>  > At L21736 ip_if.c, shouldn't the check be
 >  > 
 >  > if (!breinfo[i].bi_willdie || bireinfo[i].bi_haverep)
 >  >         continue;
 > 
 > You're right that the test at 21736 is wrong, but I think the above isn't
 > quite right either.  Instead, I think we want to identify on IREs that will
 > die and still need replacement.  Or, conversely, that we can ignore IREs
 > that (a) won't die, (b) aren't needed, or (c) are needed but have already
 > been replaced.  That would be:
 > 
 >    if (!bireinfo[i].bi_willdie || bireinfo[i].bi_haverep ||
 >        !bireinfo[i].bi_needrep)
 >            continue;

Oops -- I just went to apply this change and realized that I was thinking
about the wrong chunk of code.  I agree with your original suggestion, and
will make the change.

-- 
meem
_______________________________________________
networking-discuss mailing list
[email protected]

Reply via email to