Re: [ns] NAV bug in 802.11 MAC

["Neeraj Poojary" <legolas_3@hotmail.com>]

aman,
  the NAV fix for EIFS handling (i checked the functions set_nav(), 
reset_eifs_nav() and the call to reset_eifs_nav()) looks correct. correct me 
if i'm wrong, but it looks like you're handling the most general case where 
eifs may be different in different calls to set_nav(). if we can assume that 
the eifs is a constant (reasonable,
i think), some of the cases you're checking for will not arise.

  the original definition of eifs_ is according to the spec (sec. 9.2.10). 
so you probably shouldn't be decrementing the difs_.

  the only nitpick i have with the fix for transmission times is with the 
name DATA_Time since that macro is now used to compute the tx time of all 
frames, not just DATA frames. maybe we could leave the old ETHER_XXX_LENs 
and `XXX_Time's as they were, fix DATA_Time as you've done and use ACK_Time 
instead of DATA_Time in the initialization of eifs_...?

  i couldn't possibly diff your code against mine coz i have a bunch of 
changes too. so i just looked in the places you were likely to make fixes. 
let me know if i missed something.

-neeraj-

>From: Aman Singla <aman@atheros.com>
>To: Neeraj Poojary <legolas_3@hotmail.com>
>CC: ns-users@ISI.EDU
>Subject: Re: [ns] NAV bug in 802.11 MAC
>Date: Thu, 07 Sep 2000 15:52:31 -0700
>
>Neeraj Poojary wrote:
> >
> > aman,
> > i don't have a patch for the NAV problem. please let me know when you 
>have
> > one ready.
> >
> > you're right about the possible side-effects from creating an 
>intermediate
> > recv state. i should change my code.
> >
> > thanks,
> > neeraj
>
>I'm appending the modified files mac*802*.{cc, h}
>
>There are a bunch of other changes to the files (from the original src
>you might have) besides the above bug fix. Most of them have to do with
>using the MAC within an infrastructure BSS. Feel free to ignore them;
>I'm just too lazy to create a patch for just the above bug fix against
>the original code; anyway all my modifications shouldn't screw up
>anything
>for you.
>
>I'd greatly appreciate a code review w.r.t. the fix for the above bug;
>I'm not sure if I was really thinking straight when I did this.. and
>don't
>have a good test case.
>
>Aside: some of the calculations of transmission time were flawed w.r.t.
>the
>spec.. so I've attempted to fix them too.
>
>:a

_________________________________________________________________________
Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com.

Share information about yourself, create your own public profile at 
http://profiles.msn.com.