RE: [sv-ec] srandom, set_randstate restrictions

In my opinion, the LRM statement (at the end of 18.14.3):
 
    "An object's seed can be set from any thread. However, a thread's
seed can only be set from within the thread itself."
 
is incorrect and should be removed. 
 
 
1) Objects and threads do not have a "seed", they have a RNG state. The
RNG state is seeded (initialized.
2) Section 18.14.1 constains the bullet:
    - Manual seeding. All noninitialization RNGs can be manually seeded.
...
3) 'srandom' and 'set_randstate' are methods of the process class and
can be used to seed the
RNG of the process (thread). There should not be a restriction that the
call must be made from
the thread. Due to timing issues, it is more predictable - but not
required.
 
Regards,
Ray


________________________________

	From: owner-sv-ec@server.eda.org
[mailto:owner-sv-ec@server.eda.org] On Behalf Of Daniel Mlynek
	Sent: Tuesday, December 22, 2009 12:58 AM
	To: sv-ec@server.eda.org
	Subject: [sv-ec] srandom, set_randstate restrictions
	
	
	LRM says:
	An object's seed can be set from any thread. However, a thread's
seed can only be set from within the thread itself.
	 
	From the syntax point of view setting seed of the thread from
another thread is possible - I assume that this is semantic restriction
and below code should fail on runtime.
	My question is should LRM restriction be valid for both srandom
and set_randstate ?
	 
	 

		module top;
		 process p,p1;
		 string str;
		 initial begin
		  p = process::self();
		  #10;
		  $display($urandom());
		 end
		 
		 initial begin
		  p1 = process::self();
		  #10;
		  $display($urandom());
		 end
		 
		 initial #1 p.srandom(1); //illegal because seeding of
another thread?????
		 initial begin
		  #2;
		    str=  p.get_randstate();
		      $display(str);
		  str=  p1.get_randstate();
		  $display(str);
		   p1.set_randstate(str); //illegal because seeding of
another thread?????
		
		  end 
		endmodule
		

	 
	DANiel


	-- 
	This message has been scanned for viruses and 
	dangerous content by MailScanner <http://www.mailscanner.info/>
, and is 
	believed to be clean. 


-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.