Merge tag 'random_for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/random

Pull /dev/random changes from Ted Ts'o:
 "The /dev/random changes for 3.13 including a number of improvements in
  the following areas: performance, avoiding waste of entropy, better
  tracking of entropy estimates, support for non-x86 platforms that have
  a register which can't be used for fine-grained timekeeping, but which
  might be good enough for the random driver.

  Also add some printk's so that we can see how quickly /dev/urandom can
  get initialized, and when programs try to use /dev/urandom before it
  is fully initialized (since this could be a security issue).  This
  shouldn't be an issue on x86 desktop/laptops --- a test on my Lenovo
  T430s laptop shows that /dev/urandom is getting fully initialized
  approximately two seconds before the root file system is mounted
  read/write --- this may be an issue with ARM and MIPS embedded/mobile
  systems, though.  These printk's will be a useful canary before
  potentially adding a future change to start blocking processes which
  try to read from /dev/urandom before it is initialized, which is
  something FreeBSD does already for security reasons, and which
  security folks have been agitating for Linux to also adopt"

* tag 'random_for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/random:
  random: add debugging code to detect early use of get_random_bytes()
  random: initialize the last_time field in struct timer_rand_state
  random: don't zap entropy count in rand_initialize()
  random: printk notifications for urandom pool initialization
  random: make add_timer_randomness() fill the nonblocking pool first
  random: convert DEBUG_ENT to tracepoints
  random: push extra entropy to the output pools
  random: drop trickle mode
  random: adjust the generator polynomials in the mixing function slightly
  random: speed up the fast_mix function by a factor of four
  random: cap the rate which the /dev/urandom pool gets reseeded
  random: optimize the entropy_store structure
  random: optimize spinlock use in add_device_randomness()
  random: fix the tracepoint for get_random_bytes(_arch)
  random: account for entropy loss due to overwrites
  random: allow fractional bits to be tracked
  random: statically compute poolbitshift, poolbytes, poolbits
  random: mix in architectural randomness earlier in extract_buf()

1 files changed, 405 insertions, 242 deletions

diff --git a/drivers/char/random.c b/drivers/char/random.c
index 4fe5609eeb7..429b75bb60e 100644
--- a/drivers/char/random.c
+++ b/drivers/char/random.c
@@ -255,6 +255,7 @@
 #include <linux/fips.h>
 #include <linux/ptrace.h>
 #include <linux/kmemcheck.h>
+#include <linux/workqueue.h>
 #include <linux/irq.h>
 
 #include <asm/processor.h>
@@ -269,14 +270,28 @@
 /*
  * Configuration information
  */
-#define INPUT_POOL_WORDS 128
-#define OUTPUT_POOL_WORDS 32
-#define SEC_XFER_SIZE 512
-#define EXTRACT_SIZE 10
+#define INPUT_POOL_SHIFT	12
+#define INPUT_POOL_WORDS	(1 << (INPUT_POOL_SHIFT-5))
+#define OUTPUT_POOL_SHIFT	10
+#define OUTPUT_POOL_WORDS	(1 << (OUTPUT_POOL_SHIFT-5))
+#define SEC_XFER_SIZE		512
+#define EXTRACT_SIZE		10
+
+#define DEBUG_RANDOM_BOOT 0
 
 #define LONGS(x) (((x) + sizeof(unsigned long) - 1)/sizeof(unsigned long))
 
 /*
+ * To allow fractional bits to be tracked, the entropy_count field is
+ * denominated in units of 1/8th bits.
+ *
+ * 2*(ENTROPY_SHIFT + log2(poolbits)) must <= 31, or the multiply in
+ * credit_entropy_bits() needs to be 64 bits wide.
+ */
+#define ENTROPY_SHIFT 3
+#define ENTROPY_BITS(r) ((r)->entropy_count >> ENTROPY_SHIFT)
+
+/*
  * The minimum number of bits of entropy before we wake up a read on
  * /dev/random.  Should be enough to do a significant reseed.
  */
@@ -287,108 +302,100 @@ static int random_read_wakeup_thresh = 64;
  * should wake up processes which are selecting or polling on write
  * access to /dev/random.
  */
-static int random_write_wakeup_thresh = 128;
+static int random_write_wakeup_thresh = 28 * OUTPUT_POOL_WORDS;
 
 /*
- * When the input pool goes over trickle_thresh, start dropping most
- * samples to avoid wasting CPU time and reduce lock contention.
+ * The minimum number of seconds between urandom pool resending.  We
+ * do this to limit the amount of entropy that can be drained from the
+ * input pool even if there are heavy demands on /dev/urandom.
  */
-
-static int trickle_thresh __read_mostly = INPUT_POOL_WORDS * 28;
-
-static DEFINE_PER_CPU(int, trickle_count);
+static int random_min_urandom_seed = 60;
 
 /*
- * A pool of size .poolwords is stirred with a primitive polynomial
- * of degree .poolwords over GF(2).  The taps for various sizes are
- * defined below.  They are chosen to be evenly spaced (minimum RMS
- * distance from evenly spaced; the numbers in the comments are a
- * scaled squared error sum) except for the last tap, which is 1 to
- * get the twisting happening as fast as possible.
+ * Originally, we used a primitive polynomial of degree .poolwords
+ * over GF(2).  The taps for various sizes are defined below.  They
+ * were chosen to be evenly spaced except for the last tap, which is 1
+ * to get the twisting happening as fast as possible.
+ *
+ * For the purposes of better mixing, we use the CRC-32 polynomial as
+ * well to make a (modified) twisted Generalized Feedback Shift
+ * Register.  (See M. Matsumoto & Y. Kurita, 1992.  Twisted GFSR
+ * generators.  ACM Transactions on Modeling and Computer Simulation
+ * 2(3):179-194.  Also see M. Matsumoto & Y. Kurita, 1994.  Twisted
+ * GFSR generators II.  ACM Transactions on Mdeling and Computer
+ * Simulation 4:254-266)
+ *
+ * Thanks to Colin Plumb for suggesting this.
+ *
+ * The mixing operation is much less sensitive than the output hash,
+ * where we use SHA-1.  All that we want of mixing operation is that
+ * it be a good non-cryptographic hash; i.e. it not produce collisions
+ * when fed "random" data of the sort we expect to see.  As long as
+ * the pool state differs for different inputs, we have preserved the
+ * input entropy and done a good job.  The fact that an intelligent
+ * attacker can construct inputs that will produce controlled
+ * alterations to the pool's state is not important because we don't
+ * consider such inputs to contribute any randomness.  The only
+ * property we need with respect to them is that the attacker can't
+ * increase his/her knowledge of the pool's state.  Since all
+ * additions are reversible (knowing the final state and the input,
+ * you can reconstruct the initial state), if an attacker has any
+ * uncertainty about the initial state, he/she can only shuffle that
+ * uncertainty about, but never cause any collisions (which would
+ * decrease the uncertainty).
+ *
+ * Our mixing functions were analyzed by Lacharme, Roeck, Strubel, and
+ * Videau in their paper, "The Linux Pseudorandom Number Generator
+ * Revisited" (see: http://eprint.iacr.org/2012/251.pdf).  In their
+ * paper, they point out that we are not using a true Twisted GFSR,
+ * since Matsumoto & Kurita used a trinomial feedback polynomial (that
+ * is, with only three taps, instead of the six that we are using).
+ * As a result, the resulting polynomial is neither primitive nor
+ * irreducible, and hence does not have a maximal period over
+ * GF(2**32).  They suggest a slight change to the generator
+ * polynomial which improves the resulting TGFSR polynomial to be
+ * irreducible, which we have made here.
  */
 static struct poolinfo {
-	int poolwords;
+	int poolbitshift, poolwords, poolbytes, poolbits, poolfracbits;
+#define S(x) ilog2(x)+5, (x), (x)*4, (x)*32, (x) << (ENTROPY_SHIFT+5)
 	int tap1, tap2, tap3, tap4, tap5;
 } poolinfo_table[] = {
-	/* x^128 + x^103 + x^76 + x^51 +x^25 + x + 1 -- 105 */
-	{ 128,	103,	76,	51,	25,	1 },
-	/* x^32 + x^26 + x^20 + x^14 + x^7 + x + 1 -- 15 */
-	{ 32,	26,	20,	14,	7,	1 },
+	/* was: x^128 + x^103 + x^76 + x^51 +x^25 + x + 1 */
+	/* x^128 + x^104 + x^76 + x^51 +x^25 + x + 1 */
+	{ S(128),	104,	76,	51,	25,	1 },
+	/* was: x^32 + x^26 + x^20 + x^14 + x^7 + x + 1 */
+	/* x^32 + x^26 + x^19 + x^14 + x^7 + x + 1 */
+	{ S(32),	26,	19,	14,	7,	1 },
 #if 0
 	/* x^2048 + x^1638 + x^1231 + x^819 + x^411 + x + 1  -- 115 */
-	{ 2048,	1638,	1231,	819,	411,	1 },
+	{ S(2048),	1638,	1231,	819,	411,	1 },
 
 	/* x^1024 + x^817 + x^615 + x^412 + x^204 + x + 1 -- 290 */
-	{ 1024,	817,	615,	412,	204,	1 },
+	{ S(1024),	817,	615,	412,	204,	1 },
 
 	/* x^1024 + x^819 + x^616 + x^410 + x^207 + x^2 + 1 -- 115 */
-	{ 1024,	819,	616,	410,	207,	2 },
+	{ S(1024),	819,	616,	410,	207,	2 },
 
 	/* x^512 + x^411 + x^308 + x^208 + x^104 + x + 1 -- 225 */
-	{ 512,	411,	308,	208,	104,	1 },
+	{ S(512),	411,	308,	208,	104,	1 },
 
 	/* x^512 + x^409 + x^307 + x^206 + x^102 + x^2 + 1 -- 95 */
-	{ 512,	409,	307,	206,	102,	2 },
+	{ S(512),	409,	307,	206,	102,	2 },
 	/* x^512 + x^409 + x^309 + x^205 + x^103 + x^2 + 1 -- 95 */
-	{ 512,	409,	309,	205,	103,	2 },
+	{ S(512),	409,	309,	205,	103,	2 },
 
 	/* x^256 + x^205 + x^155 + x^101 + x^52 + x + 1 -- 125 */
-	{ 256,	205,	155,	101,	52,	1 },
+	{ S(256),	205,	155,	101,	52,	1 },
 
 	/* x^128 + x^103 + x^78 + x^51 + x^27 + x^2 + 1 -- 70 */
-	{ 128,	103,	78,	51,	27,	2 },
+	{ S(128),	103,	78,	51,	27,	2 },
 
 	/* x^64 + x^52 + x^39 + x^26 + x^14 + x + 1 -- 15 */
-	{ 64,	52,	39,	26,	14,	1 },
+	{ S(64),	52,	39,	26,	14,	1 },
 #endif
 };
 
-#define POOLBITS	poolwords*32
-#define POOLBYTES	poolwords*4
-
-/*
- * For the purposes of better mixing, we use the CRC-32 polynomial as
- * well to make a twisted Generalized Feedback Shift Reigster
- *
- * (See M. Matsumoto & Y. Kurita, 1992.  Twisted GFSR generators.  ACM
- * Transactions on Modeling and Computer Simulation 2(3):179-194.
- * Also see M. Matsumoto & Y. Kurita, 1994.  Twisted GFSR generators
- * II.  ACM Transactions on Mdeling and Computer Simulation 4:254-266)
- *
- * Thanks to Colin Plumb for suggesting this.
- *
- * We have not analyzed the resultant polynomial to prove it primitive;
- * in fact it almost certainly isn't.  Nonetheless, the irreducible factors
- * of a random large-degree polynomial over GF(2) are more than large enough
- * that periodicity is not a concern.
- *
- * The input hash is much less sensitive than the output hash.  All
- * that we want of it is that it be a good non-cryptographic hash;
- * i.e. it not produce collisions when fed "random" data of the sort
- * we expect to see.  As long as the pool state differs for different
- * inputs, we have preserved the input entropy and done a good job.
- * The fact that an intelligent attacker can construct inputs that
- * will produce controlled alterations to the pool's state is not
- * important because we don't consider such inputs to contribute any
- * randomness.  The only property we need with respect to them is that
- * the attacker can't increase his/her knowledge of the pool's state.
- * Since all additions are reversible (knowing the final state and the
- * input, you can reconstruct the initial state), if an attacker has
- * any uncertainty about the initial state, he/she can only shuffle
- * that uncertainty about, but never cause any collisions (which would
- * decrease the uncertainty).
- *
- * The chosen system lets the state of the pool be (essentially) the input
- * modulo the generator polymnomial.  Now, for random primitive polynomials,
- * this is a universal class of hash functions, meaning that the chance
- * of a collision is limited by the attacker's knowledge of the generator
- * polynomail, so if it is chosen at random, an attacker can never force
- * a collision.  Here, we use a fixed polynomial, but we *can* assume that
- * ###--> it is unknown to the processes generating the input entropy. <-###
- * Because of this important property, this is a good, collision-resistant
- * hash; hash collisions will occur no more often than chance.
- */
-
 /*
  * Static global variables
  */
@@ -396,17 +403,6 @@ static DECLARE_WAIT_QUEUE_HEAD(random_read_wait);
 static DECLARE_WAIT_QUEUE_HEAD(random_write_wait);
 static struct fasync_struct *fasync;
 
-static bool debug;
-module_param(debug, bool, 0644);
-#define DEBUG_ENT(fmt, arg...) do { \
-	if (debug) \
-		printk(KERN_DEBUG "random %04d %04d %04d: " \
-		fmt,\
-		input_pool.entropy_count,\
-		blocking_pool.entropy_count,\
-		nonblocking_pool.entropy_count,\
-		## arg); } while (0)
-
 /**********************************************************************
  *
  * OS independent entropy store.   Here are the functions which handle
@@ -417,23 +413,26 @@ module_param(debug, bool, 0644);
 struct entropy_store;
 struct entropy_store {
 	/* read-only data: */
-	struct poolinfo *poolinfo;
+	const struct poolinfo *poolinfo;
 	__u32 *pool;
 	const char *name;
 	struct entropy_store *pull;
-	int limit;
+	struct work_struct push_work;
 
 	/* read-write data: */
+	unsigned long last_pulled;
 	spinlock_t lock;
-	unsigned add_ptr;
-	unsigned input_rotate;
+	unsigned short add_ptr;
+	unsigned short input_rotate;
 	int entropy_count;
 	int entropy_total;
 	unsigned int initialized:1;
-	bool last_data_init;
+	unsigned int limit:1;
+	unsigned int last_data_init:1;
 	__u8 last_data[EXTRACT_SIZE];
 };
 
+static void push_to_pool(struct work_struct *work);
 static __u32 input_pool_data[INPUT_POOL_WORDS];
 static __u32 blocking_pool_data[OUTPUT_POOL_WORDS];
 static __u32 nonblocking_pool_data[OUTPUT_POOL_WORDS];
@@ -452,7 +451,9 @@ static struct entropy_store blocking_pool = {
 	.limit = 1,
 	.pull = &input_pool,
 	.lock = __SPIN_LOCK_UNLOCKED(blocking_pool.lock),
-	.pool = blocking_pool_data
+	.pool = blocking_pool_data,
+	.push_work = __WORK_INITIALIZER(blocking_pool.push_work,
+					push_to_pool),
 };
 
 static struct entropy_store nonblocking_pool = {
@@ -460,7 +461,9 @@ static struct entropy_store nonblocking_pool = {
 	.name = "nonblocking",
 	.pull = &input_pool,
 	.lock = __SPIN_LOCK_UNLOCKED(nonblocking_pool.lock),
-	.pool = nonblocking_pool_data
+	.pool = nonblocking_pool_data,
+	.push_work = __WORK_INITIALIZER(nonblocking_pool.push_work,
+					push_to_pool),
 };
 
 static __u32 const twist_table[8] = {
@@ -498,7 +501,7 @@ static void _mix_pool_bytes(struct entropy_store *r, const void *in,
 
 	/* mix one byte at a time to simplify size handling and churn faster */
 	while (nbytes--) {
-		w = rol32(*bytes++, input_rotate & 31);
+		w = rol32(*bytes++, input_rotate);
 		i = (i - 1) & wordmask;
 
 		/* XOR in the various taps */
@@ -518,7 +521,7 @@ static void _mix_pool_bytes(struct entropy_store *r, const void *in,
 		 * rotation, so that successive passes spread the
 		 * input bits across the pool evenly.
 		 */
-		input_rotate += i ? 7 : 14;
+		input_rotate = (input_rotate + (i ? 7 : 14)) & 31;
 	}
 
 	ACCESS_ONCE(r->input_rotate) = input_rotate;
@@ -561,65 +564,151 @@ struct fast_pool {
  * collector.  It's hardcoded for an 128 bit pool and assumes that any
  * locks that might be needed are taken by the caller.
  */
-static void fast_mix(struct fast_pool *f, const void *in, int nbytes)
+static void fast_mix(struct fast_pool *f, __u32 input[4])
 {
-	const char	*bytes = in;
 	__u32		w;
-	unsigned	i = f->count;
 	unsigned	input_rotate = f->rotate;
 
-	while (nbytes--) {
-		w = rol32(*bytes++, input_rotate & 31) ^ f->pool[i & 3] ^
-			f->pool[(i + 1) & 3];
-		f->pool[i & 3] = (w >> 3) ^ twist_table[w & 7];
-		input_rotate += (i++ & 3) ? 7 : 14;
-	}
-	f->count = i;
+	w = rol32(input[0], input_rotate) ^ f->pool[0] ^ f->pool[3];
+	f->pool[0] = (w >> 3) ^ twist_table[w & 7];
+	input_rotate = (input_rotate + 14) & 31;
+	w = rol32(input[1], input_rotate) ^ f->pool[1] ^ f->pool[0];
+	f->pool[1] = (w >> 3) ^ twist_table[w & 7];
+	input_rotate = (input_rotate + 7) & 31;
+	w = rol32(input[2], input_rotate) ^ f->pool[2] ^ f->pool[1];
+	f->pool[2] = (w >> 3) ^ twist_table[w & 7];
+	input_rotate = (input_rotate + 7) & 31;
+	w = rol32(input[3], input_rotate) ^ f->pool[3] ^ f->pool[2];
+	f->pool[3] = (w >> 3) ^ twist_table[w & 7];
+	input_rotate = (input_rotate + 7) & 31;
+
 	f->rotate = input_rotate;
+	f->count++;
 }
 
 /*
- * Credit (or debit) the entropy store with n bits of entropy
+ * Credit (or debit) the entropy store with n bits of entropy.
+ * Use credit_entropy_bits_safe() if the value comes from userspace
+ * or otherwise should be checked for extreme values.
  */
 static void credit_entropy_bits(struct entropy_store *r, int nbits)
 {
 	int entropy_count, orig;
+	const int pool_size = r->poolinfo->poolfracbits;
+	int nfrac = nbits << ENTROPY_SHIFT;
 
 	if (!nbits)
 		return;
 
-	DEBUG_ENT("added %d entropy credits to %s\n", nbits, r->name);
 retry:
 	entropy_count = orig = ACCESS_ONCE(r->entropy_count);
-	entropy_count += nbits;
+	if (nfrac < 0) {
+		/* Debit */
+		entropy_count += nfrac;
+	} else {
+		/*
+		 * Credit: we have to account for the possibility of
+		 * overwriting already present entropy.	 Even in the
+		 * ideal case of pure Shannon entropy, new contributions
+		 * approach the full value asymptotically:
+		 *
+		 * entropy <- entropy + (pool_size - entropy) *
+		 *	(1 - exp(-add_entropy/pool_size))
+		 *
+		 * For add_entropy <= pool_size/2 then
+		 * (1 - exp(-add_entropy/pool_size)) >=
+		 *    (add_entropy/pool_size)*0.7869...
+		 * so we can approximate the exponential with
+		 * 3/4*add_entropy/pool_size and still be on the
+		 * safe side by adding at most pool_size/2 at a time.
+		 *
+		 * The use of pool_size-2 in the while statement is to
+		 * prevent rounding artifacts from making the loop
+		 * arbitrarily long; this limits the loop to log2(pool_size)*2
+		 * turns no matter how large nbits is.
+		 */
+		int pnfrac = nfrac;
+		const int s = r->poolinfo->poolbitshift + ENTROPY_SHIFT + 2;
+		/* The +2 corresponds to the /4 in the denominator */
+
+		do {
+			unsigned int anfrac = min(pnfrac, pool_size/2);
+			unsigned int add =
+				((pool_size - entropy_count)*anfrac*3) >> s;
+
+			entropy_count += add;
+			pnfrac -= anfrac;
+		} while (unlikely(entropy_count < pool_size-2 && pnfrac));
+	}
 
 	if (entropy_count < 0) {
-		DEBUG_ENT("negative entropy/overflow\n");
+		pr_warn("random: negative entropy/overflow: pool %s count %d\n",
+			r->name, entropy_count);
+		WARN_ON(1);
 		entropy_count = 0;
-	} else if (entropy_count > r->poolinfo->POOLBITS)
-		entropy_count = r->poolinfo->POOLBITS;
+	} else if (entropy_count > pool_size)
+		entropy_count = pool_size;
 	if (cmpxchg(&r->entropy_count, orig, entropy_count) != orig)
 		goto retry;
 
-	if (!r->initialized && nbits > 0) {
-		r->entropy_total += nbits;
-		if (r->entropy_total > 128) {
-			r->initialized = 1;
-			if (r == &nonblocking_pool)
-				prandom_reseed_late();
+	r->entropy_total += nbits;
+	if (!r->initialized && r->entropy_total > 128) {
+		r->initialized = 1;
+		r->entropy_total = 0;
+		if (r == &nonblocking_pool) {
+			prandom_reseed_late();
+			pr_notice("random: %s pool is initialized\n", r->name);
 		}
 	}
 
-	trace_credit_entropy_bits(r->name, nbits, entropy_count,
+	trace_credit_entropy_bits(r->name, nbits,
+				  entropy_count >> ENTROPY_SHIFT,
 				  r->entropy_total, _RET_IP_);
 
-	/* should we wake readers? */
-	if (r == &input_pool && entropy_count >= random_read_wakeup_thresh) {
-		wake_up_interruptible(&random_read_wait);
-		kill_fasync(&fasync, SIGIO, POLL_IN);
+	if (r == &input_pool) {
+		int entropy_bytes = entropy_count >> ENTROPY_SHIFT;
+
+		/* should we wake readers? */
+		if (entropy_bytes >= random_read_wakeup_thresh) {
+			wake_up_interruptible(&random_read_wait);
+			kill_fasync(&fasync, SIGIO, POLL_IN);
+		}
+		/* If the input pool is getting full, send some
+		 * entropy to the two output pools, flipping back and
+		 * forth between them, until the output pools are 75%
+		 * full.
+		 */
+		if (entropy_bytes > random_write_wakeup_thresh &&
+		    r->initialized &&
+		    r->entropy_total >= 2*random_read_wakeup_thresh) {
+			static struct entropy_store *last = &blocking_pool;
+			struct entropy_store *other = &blocking_pool;
+
+			if (last == &blocking_pool)
+				other = &nonblocking_pool;
+			if (other->entropy_count <=
+			    3 * other->poolinfo->poolfracbits / 4)
+				last = other;
+			if (last->entropy_count <=
+			    3 * last->poolinfo->poolfracbits / 4) {
+				schedule_work(&last->push_work);
+				r->entropy_total = 0;
+			}
+		}
 	}
 }
 
+static void credit_entropy_bits_safe(struct entropy_store *r, int nbits)
+{
+	const int nbits_max = (int)(~0U >> (ENTROPY_SHIFT + 1));
+
+	/* Cap the value to avoid overflows */
+	nbits = min(nbits,  nbits_max);
+	nbits = max(nbits, -nbits_max);
+
+	credit_entropy_bits(r, nbits);
+}
+
 /*********************************************************************
  *
  * Entropy input management
@@ -633,6 +722,8 @@ struct timer_rand_state {
 	unsigned dont_count_entropy:1;
 };
 
+#define INIT_TIMER_RAND_STATE { INITIAL_JIFFIES, };
+
 /*
  * Add device- or boot-specific data to the input and nonblocking
  * pools to help initialize them to unique values.
@@ -644,15 +735,22 @@ struct timer_rand_state {
 void add_device_randomness(const void *buf, unsigned int size)
 {
 	unsigned long time = random_get_entropy() ^ jiffies;
+	unsigned long flags;
 
-	mix_pool_bytes(&input_pool, buf, size, NULL);
-	mix_pool_bytes(&input_pool, &time, sizeof(time), NULL);
-	mix_pool_bytes(&nonblocking_pool, buf, size, NULL);
-	mix_pool_bytes(&nonblocking_pool, &time, sizeof(time), NULL);
+	trace_add_device_randomness(size, _RET_IP_);
+	spin_lock_irqsave(&input_pool.lock, flags);
+	_mix_pool_bytes(&input_pool, buf, size, NULL);
+	_mix_pool_bytes(&input_pool, &time, sizeof(time), NULL);
+	spin_unlock_irqrestore(&input_pool.lock, flags);
+
+	spin_lock_irqsave(&nonblocking_pool.lock, flags);
+	_mix_pool_bytes(&nonblocking_pool, buf, size, NULL);
+	_mix_pool_bytes(&nonblocking_pool, &time, sizeof(time), NULL);
+	spin_unlock_irqrestore(&nonblocking_pool.lock, flags);
 }
 EXPORT_SYMBOL(add_device_randomness);
 
-static struct timer_rand_state input_timer_state;
+static struct timer_rand_state input_timer_state = INIT_TIMER_RAND_STATE;
 
 /*
  * This function adds entropy to the entropy "pool" by using timing
@@ -666,6 +764,7 @@ static struct timer_rand_state input_timer_state;
  */
 static void add_timer_randomness(struct timer_rand_state *state, unsigned num)
 {
+	struct entropy_store	*r;
 	struct {
 		long jiffies;
 		unsigned cycles;
@@ -674,15 +773,12 @@ static void add_timer_randomness(struct timer_rand_state *state, unsigned num)
 	long delta, delta2, delta3;
 
 	preempt_disable();
-	/* if over the trickle threshold, use only 1 in 4096 samples */
-	if (input_pool.entropy_count > trickle_thresh &&
-	    ((__this_cpu_inc_return(trickle_count) - 1) & 0xfff))
-		goto out;
 
 	sample.jiffies = jiffies;
 	sample.cycles = random_get_entropy();
 	sample.num = num;
-	mix_pool_bytes(&input_pool, &sample, sizeof(sample), NULL);
+	r = nonblocking_pool.initialized ? &input_pool : &nonblocking_pool;
+	mix_pool_bytes(r, &sample, sizeof(sample), NULL);
 
 	/*
 	 * Calculate number of bits of randomness we probably added.
@@ -716,10 +812,8 @@ static void add_timer_randomness(struct timer_rand_state *state, unsigned num)
 		 * Round down by 1 bit on general principles,
 		 * and limit entropy entimate to 12 bits.
 		 */
-		credit_entropy_bits(&input_pool,
-				    min_t(int, fls(delta>>1), 11));
+		credit_entropy_bits(r, min_t(int, fls(delta>>1), 11));
 	}
-out:
 	preempt_enable();
 }
 
@@ -732,10 +826,10 @@ void add_input_randomness(unsigned int type, unsigned int code,
 	if (value == last_value)
 		return;
 
-	DEBUG_ENT("input event\n");
 	last_value = value;
 	add_timer_randomness(&input_timer_state,
 			     (type << 4) ^ code ^ (code >> 4) ^ value);
+	trace_add_input_randomness(ENTROPY_BITS(&input_pool));
 }
 EXPORT_SYMBOL_GPL(add_input_randomness);
 
@@ -747,20 +841,21 @@ void add_interrupt_randomness(int irq, int irq_flags)
 	struct fast_pool	*fast_pool = &__get_cpu_var(irq_randomness);
 	struct pt_regs		*regs = get_irq_regs();
 	unsigned long		now = jiffies;
-	__u32			input[4], cycles = random_get_entropy();
-
-	input[0] = cycles ^ jiffies;
-	input[1] = irq;
-	if (regs) {
-		__u64 ip = instruction_pointer(regs);
-		input[2] = ip;
-		input[3] = ip >> 32;
-	}
+	cycles_t		cycles = random_get_entropy();
+	__u32			input[4], c_high, j_high;
+	__u64			ip;
 
-	fast_mix(fast_pool, input, sizeof(input));
+	c_high = (sizeof(cycles) > 4) ? cycles >> 32 : 0;
+	j_high = (sizeof(now) > 4) ? now >> 32 : 0;
+	input[0] = cycles ^ j_high ^ irq;
+	input[1] = now ^ c_high;
+	ip = regs ? instruction_pointer(regs) : _RET_IP_;
+	input[2] = ip;
+	input[3] = ip >> 32;
 
-	if ((fast_pool->count & 1023) &&
-	    !time_after(now, fast_pool->last + HZ))
+	fast_mix(fast_pool, input);
+
+	if ((fast_pool->count & 63) && !time_after(now, fast_pool->last + HZ))
 		return;
 
 	fast_pool->last = now;
@@ -789,10 +884,8 @@ void add_disk_randomness(struct gendisk *disk)
 	if (!disk || !disk->random)
 		return;
 	/* first major is 1, so we get >= 0x200 here */
-	DEBUG_ENT("disk event %d:%d\n",
-		  MAJOR(disk_devt(disk)), MINOR(disk_devt(disk)));
-
 	add_timer_randomness(disk->random, 0x100 + disk_devt(disk));
+	trace_add_disk_randomness(disk_devt(disk), ENTROPY_BITS(&input_pool));
 }
 #endif
 
@@ -810,30 +903,58 @@ static ssize_t extract_entropy(struct entropy_store *r, void *buf,
  * from the primary pool to the secondary extraction pool. We make
  * sure we pull enough for a 'catastrophic reseed'.
  */
+static void _xfer_secondary_pool(struct entropy_store *r, size_t nbytes);
 static void xfer_secondary_pool(struct entropy_store *r, size_t nbytes)
 {
-	__u32	tmp[OUTPUT_POOL_WORDS];
+	if (r->limit == 0 && random_min_urandom_seed) {
+		unsigned long now = jiffies;
 
-	if (r->pull && r->entropy_count < nbytes * 8 &&
-	    r->entropy_count < r->poolinfo->POOLBITS) {
-		/* If we're limited, always leave two wakeup worth's BITS */
-		int rsvd = r->limit ? 0 : random_read_wakeup_thresh/4;
-		int bytes = nbytes;
-
-		/* pull at least as many as BYTES as wakeup BITS */
-		bytes = max_t(int, bytes, random_read_wakeup_thresh / 8);
-		/* but never more than the buffer size */
-		bytes = min_t(int, bytes, sizeof(tmp));
-
-		DEBUG_ENT("going to reseed %s with %d bits "
-			  "(%zu of %d requested)\n",
-			  r->name, bytes * 8, nbytes * 8, r->entropy_count);
-
-		bytes = extract_entropy(r->pull, tmp, bytes,
-					random_read_wakeup_thresh / 8, rsvd);
-		mix_pool_bytes(r, tmp, bytes, NULL);
-		credit_entropy_bits(r, bytes*8);
+		if (time_before(now,
+				r->last_pulled + random_min_urandom_seed * HZ))
+			return;
+		r->last_pulled = now;
 	}
+	if (r->pull &&
+	    r->entropy_count < (nbytes << (ENTROPY_SHIFT + 3)) &&
+	    r->entropy_count < r->poolinfo->poolfracbits)
+		_xfer_secondary_pool(r, nbytes);
+}
+
+static void _xfer_secondary_pool(struct entropy_store *r, size_t nbytes)
+{
+	__u32	tmp[OUTPUT_POOL_WORDS];
+
+	/* For /dev/random's pool, always leave two wakeup worth's BITS */
+	int rsvd = r->limit ? 0 : random_read_wakeup_thresh/4;
+	int bytes = nbytes;
+
+	/* pull at least as many as BYTES as wakeup BITS */
+	bytes = max_t(int, bytes, random_read_wakeup_thresh / 8);
+	/* but never more than the buffer size */
+	bytes = min_t(int, bytes, sizeof(tmp));
+
+	trace_xfer_secondary_pool(r->name, bytes * 8, nbytes * 8,
+				  ENTROPY_BITS(r), ENTROPY_BITS(r->pull));
+	bytes = extract_entropy(r->pull, tmp, bytes,
+				random_read_wakeup_thresh / 8, rsvd);
+	mix_pool_bytes(r, tmp, bytes, NULL);
+	credit_entropy_bits(r, bytes*8);
+}
+
+/*
+ * Used as a workqueue function so that when the input pool is getting
+ * full, we can "spill over" some entropy to the output pools.  That
+ * way the output pools can store some of the excess entropy instead
+ * of letting it go to waste.
+ */
+static void push_to_pool(struct work_struct *work)
+{
+	struct entropy_store *r = container_of(work, struct entropy_store,
+					      push_work);
+	BUG_ON(!r);
+	_xfer_secondary_pool(r, random_read_wakeup_thresh/8);
+	trace_push_to_pool(r->name, r->entropy_count >> ENTROPY_SHIFT,
+			   r->pull->entropy_count >> ENTROPY_SHIFT);
 }
 
 /*
@@ -853,50 +974,48 @@ static size_t account(struct entropy_store *r, size_t nbytes, int min,
 {
 	unsigned long flags;
 	int wakeup_write = 0;
+	int have_bytes;
+	int entropy_count, orig;
+	size_t ibytes;
 
 	/* Hold lock while accounting */
 	spin_lock_irqsave(&r->lock, flags);
 
-	BUG_ON(r->entropy_count > r->poolinfo->POOLBITS);
-	DEBUG_ENT("trying to extract %zu bits from %s\n",
-		  nbytes * 8, r->name);
+	BUG_ON(r->entropy_count > r->poolinfo->poolfracbits);
 
 	/* Can we pull enough? */
-	if (r->entropy_count / 8 < min + reserved) {
-		nbytes = 0;
-	} else {
-		int entropy_count, orig;
 retry:
-		entropy_count = orig = ACCESS_ONCE(r->entropy_count);
+	entropy_count = orig = ACCESS_ONCE(r->entropy_count);
+	have_bytes = entropy_count >> (ENTROPY_SHIFT + 3);
+	ibytes = nbytes;
+	if (have_bytes < min + reserved) {
+		ibytes = 0;
+	} else {
 		/* If limited, never pull more than available */
-		if (r->limit && nbytes + reserved >= entropy_count / 8)
-			nbytes = entropy_count/8 - reserved;
-
-		if (entropy_count / 8 >= nbytes + reserved) {
-			entropy_count -= nbytes*8;
-			if (cmpxchg(&r->entropy_count, orig, entropy_count) != orig)
-				goto retry;
-		} else {
-			entropy_count = reserved;
-			if (cmpxchg(&r->entropy_count, orig, entropy_count) != orig)
-				goto retry;
-		}
+		if (r->limit && ibytes + reserved >= have_bytes)
+			ibytes = have_bytes - reserved;
 
-		if (entropy_count < random_write_wakeup_thresh)
-			wakeup_write = 1;
-	}
+		if (have_bytes >= ibytes + reserved)
+			entropy_count -= ibytes << (ENTROPY_SHIFT + 3);
+		else
+			entropy_count = reserved << (ENTROPY_SHIFT + 3);
 
-	DEBUG_ENT("debiting %zu entropy credits from %s%s\n",
-		  nbytes * 8, r->name, r->limit ? "" : " (unlimited)");
+		if (cmpxchg(&r->entropy_count, orig, entropy_count) != orig)
+			goto retry;
 
+		if ((r->entropy_count >> ENTROPY_SHIFT)
+		    < random_write_wakeup_thresh)
+			wakeup_write = 1;
+	}
 	spin_unlock_irqrestore(&r->lock, flags);
 
+	trace_debit_entropy(r->name, 8 * ibytes);
 	if (wakeup_write) {
 		wake_up_interruptible(&random_write_wait);
 		kill_fasync(&fasync, SIGIO, POLL_OUT);
 	}
 
-	return nbytes;
+	return ibytes;
 }
 
 static void extract_buf(struct entropy_store *r, __u8 *out)
@@ -904,7 +1023,7 @@ static void extract_buf(struct entropy_store *r, __u8 *out)
 	int i;
 	union {
 		__u32 w[5];
-		unsigned long l[LONGS(EXTRACT_SIZE)];
+		unsigned long l[LONGS(20)];
 	} hash;
 	__u32 workspace[SHA_WORKSPACE_WORDS];
 	__u8 extract[64];
@@ -917,6 +1036,17 @@ static void extract_buf(struct entropy_store *r, __u8 *out)
 		sha_transform(hash.w, (__u8 *)(r->pool + i), workspace);
 
 	/*
+	 * If we have a architectural hardware random number
+	 * generator, mix that in, too.
+	 */
+	for (i = 0; i < LONGS(20); i++) {
+		unsigned long v;
+		if (!arch_get_random_long(&v))
+			break;
+		hash.l[i] ^= v;
+	}
+
+	/*
 	 * We mix the hash back into the pool to prevent backtracking
 	 * attacks (where the attacker knows the state of the pool
 	 * plus the current outputs, and attempts to find previous
@@ -945,17 +1075,6 @@ static void extract_buf(struct entropy_store *r, __u8 *out)
 	hash.w[1] ^= hash.w[4];
 	hash.w[2] ^= rol32(hash.w[2], 16);
 
-	/*
-	 * If we have a architectural hardware random number
-	 * generator, mix that in, too.
-	 */
-	for (i = 0; i < LONGS(EXTRACT_SIZE); i++) {
-		unsigned long v;
-		if (!arch_get_random_long(&v))
-			break;
-		hash.l[i] ^= v;
-	}
-
 	memcpy(out, &hash, EXTRACT_SIZE);
 	memset(&hash, 0, sizeof(hash));
 }
@@ -971,10 +1090,10 @@ static ssize_t extract_entropy(struct entropy_store *r, void *buf,
 	if (fips_enabled) {
 		spin_lock_irqsave(&r->lock, flags);
 		if (!r->last_data_init) {
-			r->last_data_init = true;
+			r->last_data_init = 1;
 			spin_unlock_irqrestore(&r->lock, flags);
 			trace_extract_entropy(r->name, EXTRACT_SIZE,
-					      r->entropy_count, _RET_IP_);
+					      ENTROPY_BITS(r), _RET_IP_);
 			xfer_secondary_pool(r, EXTRACT_SIZE);
 			extract_buf(r, tmp);
 			spin_lock_irqsave(&r->lock, flags);
@@ -983,7 +1102,7 @@ static ssize_t extract_entropy(struct entropy_store *r, void *buf,
 		spin_unlock_irqrestore(&r->lock, flags);
 	}
 
-	trace_extract_entropy(r->name, nbytes, r->entropy_count, _RET_IP_);
+	trace_extract_entropy(r->name, nbytes, ENTROPY_BITS(r), _RET_IP_);
 	xfer_secondary_pool(r, nbytes);
 	nbytes = account(r, nbytes, min, reserved);
 
@@ -1016,7 +1135,7 @@ static ssize_t extract_entropy_user(struct entropy_store *r, void __user *buf,
 	ssize_t ret = 0, i;
 	__u8 tmp[EXTRACT_SIZE];
 
-	trace_extract_entropy_user(r->name, nbytes, r->entropy_count, _RET_IP_);
+	trace_extract_entropy_user(r->name, nbytes, ENTROPY_BITS(r), _RET_IP_);
 	xfer_secondary_pool(r, nbytes);
 	nbytes = account(r, nbytes, 0, 0);
 
@@ -1056,6 +1175,14 @@ static ssize_t extract_entropy_user(struct entropy_store *r, void __user *buf,
  */
 void get_random_bytes(void *buf, int nbytes)
 {
+#if DEBUG_RANDOM_BOOT > 0
+	if (unlikely(nonblocking_pool.initialized == 0))
+		printk(KERN_NOTICE "random: %pF get_random_bytes called "
+		       "with %d bits of entropy available\n",
+		       (void *) _RET_IP_,
+		       nonblocking_pool.entropy_total);
+#endif
+	trace_get_random_bytes(nbytes, _RET_IP_);
 	extract_entropy(&nonblocking_pool, buf, nbytes, 0, 0);
 }
 EXPORT_SYMBOL(get_random_bytes);
@@ -1074,7 +1201,7 @@ void get_random_bytes_arch(void *buf, int nbytes)
 {
 	char *p = buf;
 
-	trace_get_random_bytes(nbytes, _RET_IP_);
+	trace_get_random_bytes_arch(nbytes, _RET_IP_);
 	while (nbytes) {
 		unsigned long v;
 		int chunk = min(nbytes, (int)sizeof(unsigned long));
@@ -1108,13 +1235,11 @@ static void init_std_data(struct entropy_store *r)
 	ktime_t now = ktime_get_real();
 	unsigned long rv;
 
-	r->entropy_count = 0;
-	r->entropy_total = 0;
-	r->last_data_init = false;
+	r->last_pulled = jiffies;
 	mix_pool_bytes(r, &now, sizeof(now), NULL);
-	for (i = r->poolinfo->POOLBYTES; i > 0; i -= sizeof(rv)) {
+	for (i = r->poolinfo->poolbytes; i > 0; i -= sizeof(rv)) {
 		if (!arch_get_random_long(&rv))
-			break;
+			rv = random_get_entropy();
 		mix_pool_bytes(r, &rv, sizeof(rv), NULL);
 	}
 	mix_pool_bytes(r, utsname(), sizeof(*(utsname())), NULL);
@@ -1137,7 +1262,7 @@ static int rand_initialize(void)
 	init_std_data(&nonblocking_pool);
 	return 0;
 }
-module_init(rand_initialize);
+early_initcall(rand_initialize);
 
 #ifdef CONFIG_BLOCK
 void rand_initialize_disk(struct gendisk *disk)
@@ -1149,8 +1274,10 @@ void rand_initialize_disk(struct gendisk *disk)
 	 * source.
 	 */
 	state = kzalloc(sizeof(struct timer_rand_state), GFP_KERNEL);
-	if (state)
+	if (state) {
+		state->last_time = INITIAL_JIFFIES;
 		disk->random = state;
+	}
 }
 #endif
 
@@ -1167,8 +1294,6 @@ random_read(struct file *file, char __user *buf, size_t nbytes, loff_t *ppos)
 		if (n > SEC_XFER_SIZE)
 			n = SEC_XFER_SIZE;
 
-		DEBUG_ENT("reading %zu bits\n", n*8);
-
 		n = extract_entropy_user(&blocking_pool, buf, n);
 
 		if (n < 0) {
@@ -1176,8 +1301,9 @@ random_read(struct file *file, char __user *buf, size_t nbytes, loff_t *ppos)
 			break;
 		}
 
-		DEBUG_ENT("read got %zd bits (%zd still needed)\n",
-			  n*8, (nbytes-n)*8);
+		trace_random_read(n*8, (nbytes-n)*8,
+				  ENTROPY_BITS(&blocking_pool),
+				  ENTROPY_BITS(&input_pool));
 
 		if (n == 0) {
 			if (file->f_flags & O_NONBLOCK) {
@@ -1185,13 +1311,9 @@ random_read(struct file *file, char __user *buf, size_t nbytes, loff_t *ppos)
 				break;
 			}
 
-			DEBUG_ENT("sleeping?\n");
-
 			wait_event_interruptible(random_read_wait,
-				input_pool.entropy_count >=
-						 random_read_wakeup_thresh);
-
-			DEBUG_ENT("awake\n");
+				ENTROPY_BITS(&input_pool) >=
+				random_read_wakeup_thresh);
 
 			if (signal_pending(current)) {
 				retval = -ERESTARTSYS;
@@ -1214,7 +1336,18 @@ random_read(struct file *file, char __user *buf, size_t nbytes, loff_t *ppos)
 static ssize_t
 urandom_read(struct file *file, char __user *buf, size_t nbytes, loff_t *ppos)
 {
-	return extract_entropy_user(&nonblocking_pool, buf, nbytes);
+	int ret;
+
+	if (unlikely(nonblocking_pool.initialized == 0))
+		printk_once(KERN_NOTICE "random: %s urandom read "
+			    "with %d bits of entropy available\n",
+			    current->comm, nonblocking_pool.entropy_total);
+
+	ret = extract_entropy_user(&nonblocking_pool, buf, nbytes);
+
+	trace_urandom_read(8 * nbytes, ENTROPY_BITS(&nonblocking_pool),
+			   ENTROPY_BITS(&input_pool));
+	return ret;
 }
 
 static unsigned int
@@ -1225,9 +1358,9 @@ random_poll(struct file *file, poll_table * wait)
 	poll_wait(file, &random_read_wait, wait);
 	poll_wait(file, &random_write_wait, wait);
 	mask = 0;
-	if (input_pool.entropy_count >= random_read_wakeup_thresh)
+	if (ENTROPY_BITS(&input_pool) >= random_read_wakeup_thresh)
 		mask |= POLLIN | POLLRDNORM;
-	if (input_pool.entropy_count < random_write_wakeup_thresh)
+	if (ENTROPY_BITS(&input_pool) < random_write_wakeup_thresh)
 		mask |= POLLOUT | POLLWRNORM;
 	return mask;
 }
@@ -1278,7 +1411,8 @@ static long random_ioctl(struct file *f, unsigned int cmd, unsigned long arg)
 	switch (cmd) {
 	case RNDGETENTCNT:
 		/* inherently racy, no point locking */
-		if (put_user(input_pool.entropy_count, p))
+		ent_count = ENTROPY_BITS(&input_pool);
+		if (put_user(ent_count, p))
 			return -EFAULT;
 		return 0;
 	case RNDADDTOENTCNT:
@@ -1286,7 +1420,7 @@ static long random_ioctl(struct file *f, unsigned int cmd, unsigned long arg)
 			return -EPERM;
 		if (get_user(ent_count, p))
 			return -EFAULT;
-		credit_entropy_bits(&input_pool, ent_count);
+		credit_entropy_bits_safe(&input_pool, ent_count);
 		return 0;
 	case RNDADDENTROPY:
 		if (!capable(CAP_SYS_ADMIN))
@@ -1301,14 +1435,19 @@ static long random_ioctl(struct file *f, unsigned int cmd, unsigned long arg)
 				    size);
 		if (retval < 0)
 			return retval;
-		credit_entropy_bits(&input_pool, ent_count);
+		credit_entropy_bits_safe(&input_pool, ent_count);
 		return 0;
 	case RNDZAPENTCNT:
 	case RNDCLEARPOOL:
-		/* Clear the entropy pool counters. */
+		/*
+		 * Clear the entropy pool counters. We no longer clear
+		 * the entropy pool, as that's silly.
+		 */
 		if (!capable(CAP_SYS_ADMIN))
 			return -EPERM;
-		rand_initialize();
+		input_pool.entropy_count = 0;
+		nonblocking_pool.entropy_count = 0;
+		blocking_pool.entropy_count = 0;
 		return 0;
 	default:
 		return -EINVAL;
@@ -1408,6 +1547,23 @@ static int proc_do_uuid(struct ctl_table *table, int write,
 	return proc_dostring(&fake_table, write, buffer, lenp, ppos);
 }
 
+/*
+ * Return entropy available scaled to integral bits
+ */
+static int proc_do_entropy(ctl_table *table, int write,
+			   void __user *buffer, size_t *lenp, loff_t *ppos)
+{
+	ctl_table fake_table;
+	int entropy_count;
+
+	entropy_count = *(int *)table->data >> ENTROPY_SHIFT;
+
+	fake_table.data = &entropy_count;
+	fake_table.maxlen = sizeof(entropy_count);
+
+	return proc_dointvec(&fake_table, write, buffer, lenp, ppos);
+}
+
 static int sysctl_poolsize = INPUT_POOL_WORDS * 32;
 extern struct ctl_table random_table[];
 struct ctl_table random_table[] = {
@@ -1422,7 +1578,7 @@ struct ctl_table random_table[] = {
 		.procname	= "entropy_avail",
 		.maxlen		= sizeof(int),
 		.mode		= 0444,
-		.proc_handler	= proc_dointvec,
+		.proc_handler	= proc_do_entropy,
 		.data		= &input_pool.entropy_count,
 	},
 	{
@@ -1444,6 +1600,13 @@ struct ctl_table random_table[] = {
 		.extra2		= &max_write_thresh,
 	},
 	{
+		.procname	= "urandom_min_reseed_secs",
+		.data		= &random_min_urandom_seed,
+		.maxlen		= sizeof(int),
+		.mode		= 0644,
+		.proc_handler	= proc_dointvec,
+	},
+	{
 		.procname	= "boot_id",
 		.data		= &sysctl_bootid,
 		.maxlen		= 16,