[RC5] LIFO? Is this true?

gindrup at okway.okstate.edu gindrup at okway.okstate.edu
Wed May 13 11:21:33 EDT 1998


     Actually, on the platforms in question, either there is adequate 
     record-level locking, or there is not.  If there is, the lock bits 
     are superfluous as the locking structure is owned and managed by the 
     system hosting the buff-* files.  If there is not, then I am under 
     the impression from previous group postings that either the platform 
     has sufficient page-level locking to get the job done (MS OSes) or 
     the client has not been ported to the platform.  Either way the bits 
     still aren't needed.
     
     
     One local person has suggested an alternative that appeals to me 
     somewhat.  File Chaining.  Each fetched batch of blocks is in a new 
     file (one file per batch, *not* one file per block).  Each file 
     contains a filespec to the file that follows it.  Blocks are removed 
     from a given file until it is exhausted.  Subsequent blocks come 
     from the file specified in the "next file" field in the file header. 
      (Oh no, a data structure!  :-)  The last client to finish a block 
     in the buff-in file deletes it and renames the next file in the 
     chain "buff-in".
     
     This can get slow if there are *many* clients and there are frequent 
     fetches.  Unless the cracking threads get to where they are cracking 
     entire blocks in a couple of seconds, though, this won't affect the 
     mt clients at all...
     
     


______________________________ Reply Separator _________________________________
Subject: Re: Re[2]: [RC5] LIFO?  Is this true?  
Author:  <rc5 at llamas.net> at SMTP
Date:    5/12/98 3:23 PM


     [snip]
>      Clients "should" do record-level locking.  
[snip = but MS OSes can't without help]
>      Finally, a FIFO buff-* file doesn't require any data structure.  
.. Except a two-bit-per-block lock array (one bit for read, one for 
write.  I don't think we ever need to block one and not the other, but 
just in case it isn't much extra space).
     [snip]
     
        -=- James Mastros
-- 
"True mastery is knowing enough to bullshit the rest."
        -=- Me

--
To unsubscribe, send 'unsubscribe rc5' to majordomo at lists.distributed.net
rc5-digest subscribers replace rc5 with rc5-digest



More information about the rc5 mailing list