Re: [patch] Re: Can modules block on socket writes? (fwd)

["David M. Grimes" <dmgrime appliedtheory com>]

On Tue, Mar 13, 2001 at 09:51:30PM +0100, Marcus Gruendler wrote:
> On Tuesday 13 March 2001 20:34 you wrote:
> >
> > does this mechanism solve your problem?
> >
> 
> I just had another _really nice_ (tm) idea! What do you think about a 
> SEND_BUFFER action that awaits a vector of (pointer to) buffers. This would 
> enable a scatter gather write behaviour just like writev(). What do you think 
> about that?


That would be nice - I don't know how much of a leap it is from the
SEND_BUFFER that is there as of -S6, Ingo?  BTW - I understand the
SEND_BUFFER to take the user buffer as input and manage any fragmentation
internally - just rescheduling the user_req_t when the buffer has been fully
sent, or removing it if it was closed?  Is there a way to notify userspace
if the send failed?

Thanks!

  Dave
  
> 
> Bye, Marcus 
> 
> -- 
> Marcus Gruendler
> eMail: runner@tamalin.de
> WWW: http://www.tamalin.de/runner/index.html
> 
> 
> 
> _______________________________________________
> tux-list mailing list
> tux-list@redhat.com
> https://listman.redhat.com/mailman/listinfo/tux-list