Subject: Re: PROPOSAL: removal of brk()/sbrk().
To: None <tech-kern@netbsd.org, tech-userlevel@netbsd.org>
From: der Mouse <mouse@Rodents.Montreal.QC.CA>
List: tech-kern
Date: 02/25/2002 18:20:05
>> What about legacy static binaries [that use brk()]?
>> Or Am I Missing Something?
> Isn't that covered with
> 	Set sys_obreak() in kernel #ifdef COMPAT_15 and change start
> 	of mmap space calculation. Quite straightforward.

Maybe, depending on what that sentence actually refers to.  I didn't
see any actual code for that part of it, though admittedly I may have
searched for the wrong thing.

If that sentence means that kernels with COMPAT_15 (and earlier!) use
basically the status quo (or perhaps implementing one of the methods
that reserves only max-stack-size rather than max-data-size VM
space)...in that case, yes, that covers it, but in that case I'm not
entirely clear what the perceived benefit is.  Will we be deprecating
COMPAT_ options for 1.5 and earlier "soon"?  I don't see any other way
to make it work; a legacy binary could very well do an mmap before
first trying to brk, which makes it hard to tell whether it's a legacy
binary at mmap time.

/~\ The ASCII				der Mouse
\ / Ribbon Campaign
 X  Against HTML	       mouse@rodents.montreal.qc.ca
/ \ Email!	     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B