Date: Fri, 16 May 1997 10:27:14 -0400 (EDT) From: Andrew Gallatin <gallatin@CS.Duke.EDU> To: "Chris G. Demetriou" <cgd@pa.dec.com> Cc: Warner Losh <imp@village.org>, freebsd-alpha@FreeBSD.ORG Subject: Re: Alpha questions.. Message-ID: <199705161427.KAA05260@hurricane.cs.duke.edu> In-Reply-To: <717.863741892@dnaunix.pa.dec.com> References: <199705152232.SAA28646@hurricane.cs.duke.edu> <717.863741892@dnaunix.pa.dec.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Chris G. Demetriou writes: > > > : Also, I understand that the PALcode in MILO is not idential to the > > > : PALcode in the SRM console firmware. > > > > Download milo & have a look. The palcode claims to be the OSF > > PAL code. > > It is, but it is _not_ identical to the OSF PALcode in the SRM console > firmware in several ways. > > At minimum: > > (1) there are some interfaces provided by the SRM console's > OSF PALcode on some platforms which are not provided > by the MILO OSF PALcode, and > > (2) machine-checking handling/logout frames are different. Thanks for clearing that up. > > Also, since Linux can boot from either SRM or Milo, it would seem that > > the OSF PAL code in Milo & in the SRM console are "close enough." > > They are not. > > There are operations provided by the OSF PALcode on (at least) EB64+ > and EB164 machines which are not provided by MILO's OSF PALcode. > > If you're using EB64+ or EB164 systems with 'native' OSF PALcode and > try to use the hardware directly rather than using the PALcode ops > provided by the firmware (for the appropriate operations), you will > not be receiving many interrupts... It looks like linux uses conditional compilation to deal with those issues, thus elminating the possibility of a truley generic kernel. Ick. Thanks again, Drew
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199705161427.KAA05260>