lkml.org 
[lkml]   [1998]   [Sep]   [20]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: Linux, UDI and SCO.
On Sun, 20 Sep 1998, Michael Meissner wrote:

> You know this UDI debate is amusing, coming about one month after there was
> lots of moaning about Linus breaking drivers due to adding a field in the
> middle of driver vector table, forcing new versions of the externally supplied
> drivers (like PCMCIA for instance). If you go with UDI, you are essentially
> tying the hands of the developers to adhere to it, since you don't even have
> the choice of a recompile.
>
> This is similar to a problem in userspace land -- on the x86, we use the System
> V.4 ABI for setting things up. The V.4 ABI was created in the days of 386's,
> and was demands that doubles have *4* byte alignment, not the more natural *8*
> byte alignment (and the stack is similarly aligned to a *4* byte boundary).
> Ever since the Pentium came out, we compiler folk have been asked to make the
> default be the better alignment, but we can't, since we would then be breaking
> the ABI (and hence binary compatibility with objects compiled with the old
> compiler).

It would be nice if we could declare the PII to be a differnt arch and
make it incompatible with anything pre-pentium. But since this would be an
atrifical limitation, users would scream that the hardware supports it so
we should to..

The day mainstream x86 dies will be joyous, compatiblity is SUCH a
burden.. Every arch needs die out after 10 years..



-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2005-03-22 13:44    [from the cache]
©2003-2011 Jasper Spaans