[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: gnumach ChangeLog i386/i386/locore.h i386/i386a... [gnumach-1-branch
From: |
Samuel Thibault |
Subject: |
Re: gnumach ChangeLog i386/i386/locore.h i386/i386a... [gnumach-1-branch] |
Date: |
Thu, 24 Jul 2008 11:51:17 +0100 |
User-agent: |
Mutt/1.5.12-2006-07-14 |
Thomas Schwinge, le Thu 24 Jul 2008 12:44:42 +0200, a écrit :
> extern int copyin (const void *userbuf, void *kernelbuf, size_t cn);
>
> -extern int copyinmsg (vm_offset_t userbuf, vm_offset_t kernelbuf, size_t
> cn);
> +extern int copyinmsg (const void *userbuf, void *kernelbuf, size_t cn);
>
> See `i386/i386/locore.S': `copyin' and `copyinmsg' (and same for the out
> variants) already have the same entry points. Now you also made them
> equivalent at the C language level. Perhaps just get rid of one of those
> two symbols? Or was there a real reason to have both functions with
> different C language signatures point to the same implementation?
Oh, how odd. The only reason I could see would be to have
copyin/copyinmsg differ just by one taking pointers and the other taking
integers, but what was there before Barry's fixes was rather the
converse (i.e. copyinmsg is used on buffers). I guess we should dig
back further in the history of Mach, there should be a reason which
might bite use in the future.
Samuel