On 18-08-08 19:25, Stanichenko Marat wrote:

quoted text
> From: Stanichenko Marat <mstanichenko@gmail.com>
> 
> Add the interface that shows the amount of entropy generated via the 
> interfaces and also the amount consumed. This patch adds two files in 
> proc. The first of them shows the entropy generated per interface, the 
> second one - the amount consumed from blocking (/dev/random)  and 
> nonblocking (/dev/urandom) pools.
> 
> This patch is an attempt to realize "/dev/random statistics" project of 
> kernel newbies community.
> http://kernelnewbies.org/KernelProjects
> 
> Signed-off-by: Stanichenko Marat <mstanichenko@gmail.com>

Just two quick comments/questions as encountered while reading. I'm not 
familiar with the random code or anything.

First, your patch was posted Base64 encoded which doesn't make it easier 
to look at/comment on. Base64 encoding even text/plain attchments is a 
longstanding Thunderbird bug when your outgoing charset is set to UTF-8 
(and the reason why my outgoing charset is ISO8859-15; you just won't 
remember to switch from UTF-8 to something else every time when posting 
a patch otherwise).

quoted text
> +struct ioctl_rand_stat {
> +	unsigned nbits;
> +	spinlock_t lock;
> +};
> +static struct ioctl_rand_stat  ent_ioctl_produce = {
> +	.nbits = 0,
> +	.lock = __SPIN_LOCK_UNLOCKED(&input_timer_state.lock),

Is that supposed to be &ent_ioctl_produce.lock?

quoted text
> @@ -621,8 +645,11 @@ static void add_timer_randomness(struct 
>  		 * 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));
> +		nbits = min_t(int, fls(delta>>1), 11);
> +		spin_lock_irqsave(&state->lock, flags);
> +		state->nbits += nbits;
> +		spin_unlock_irqrestore(&state->lock, flags);
> +		credit_entropy_bits(&input_pool,nbits);


Couldn't state->nbits be an atomic_t? The locking looks like it might be 
a little expensive. Maybe they should be per-cpu? I have no idea about 
the frequencies here...

Rene.

--
To unsubscribe from this list: send an email with
"unsubscribe kernelnewbies" to ecartis@nl.linux.org
Please read the FAQ at http://kernelnewbies.org/FAQ