ef45cc76   
   From: rick.jones2@hp.com   
      
   David Schwartz  wrote:   
   > On Mar 1, 1:02?pm, Rick Jones  wrote:   
      
   > > I do recall that *generating* a UDP checksum is optional, I don't   
   > > recall the specification saying that validating it was optional.   
      
   > That was my recollection, but I could be incorrect. But it really   
   > doesn't affect any of the major issues here. The Linux behavior is   
   > still legal because UDP datagrams are discardable.   
      
   Until Linux's behaviour came around the ass-u-me-ption was that the   
   discardability was at layers 4 and below, not between UDP and the   
   socket layer.   
      
   UDP datagrams being discardable by the lower layers was used as cover   
   to say that the semantics of copy-and-checksum-oops and/or   
   now-we-want-those-buffers-back were "OK" and it was not necessary to   
   leave some sort of tombstone behind to assure the next access of the   
   socket would have something to do, even it it was to report some sort   
   of transient error.   
      
   (FWIW, I think that was the motivation behind the odd "ENOBUFS" return   
   of accept() in HP-UX 11 - because all those pesky enterprise customers   
   who were accustomed to a particular behaviour - not getting stuck in   
   an accept() call - wanted that maintained) HP-UX 11 was an "upgrade"   
   to previous HP-UX versions and so backwards compatability to   
   semantics, even those in grey areas of what few standards there were   
   were rather important).   
      
   rick jones   
   --   
   a wide gulf separates "what if" from "if only"   
   these opinions are mine, all mine; HP might not want them anyway... :)   
   feel free to post, OR email to rick.jones2 in hp.com but NOT BOTH...   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)