> >>>> i have an implementation of lwp-specific data.
> >> (attached.  i wrote it to replace l_emuldata, l_savp, etc.)
> >> proc-specific data can be similar except locking.
> >
> > Ok, I hereby volunteer to generalize this and make it something that  
> > we can use with procs, lwps, and possibly sessions (I can see some  
> > value in extensible per-session data as well).
> >
> > I'll work on it this week.
> 
> Ok, here is a generalized (and currently untested) "specificdata"  
> implementation, based on Yamamoto-san's.  proc, lwp, and session  
> specific-data could be built on top of this, a'la (note this is  
> incomplete):

thanks for working on it.

> Comments?

my original intention was that, an lwp only manipulates
its own lwp specific data.  (except special cases like lwp-startup)
by limiting the usage in this way, getspecific can be very cheap.
ie. no locking or memory-writes.
i'm not sure how much difference it can make, tho.

YAMAMOTO Takashi