> On Nov 9, 12:42pm, yamt@mwd.biglobe.ne.jp (YAMAMOTO Takashi) wrote:
> -- Subject: Re: CVS commit: src/lib/libc/gen
> 
> | > | doesn't it break vfork+execvp?
> | > 
> | > good point. I guess I can use mmap.
> | > 
> | > christos
> | 
> | how does mmap solve it?
> 
> It can be undone safely...
> 
> christos

well, i'm not sure if we are talking about the same problem.

when execve succeeds, code after that will not be executed
whehter it's free or munmap.  in the case of vfork+exec,
it means memory leak on the parent process.

YAMAMOTO Takashi