Date: Thu, 13 Dec 2001 16:28:50 -0800 From: Mike Smith <msmith@freebsd.org> To: Thomas Moestl <tmoestl@gmx.net> Cc: Terry Lambert <tlambert2@mindspring.com>, arch@FreeBSD.org Subject: Re: Please review: changes to MI bus code for sparc64 Message-ID: <200112140028.fBE0Sol04630@mass.dis.org> In-Reply-To: Message from Thomas Moestl <tmoestl@gmx.net> of "Thu, 13 Dec 2001 20:12:13 %2B0100." <20011213201213.B871@crow.dom2ip.de>
next in thread | previous in thread | raw e-mail | index | archive | help
> > The PCI_BROKEN_INTPIN/PCI_INTLINE_0_BAD seem to be the same thing; > > they should be protected by a single name (probably PCI_BROKEN_INTPIN) > > in the #ifdef in pci.c; it should be "all or nothing" on a single > > value. As it is, you must define one if you define the other, but > > not vice versa, and the effect seems to be linked, anyway, so you > > might as well use a single protection mechanism. > > It is not uncommon that i386 BIOSes to set the intline register to 0 > when it should really be 0xff (to indicate an unrouted interrupt). So, > I figured that it might be useful to make this an extra option. No. Fix the i386-specific config space accessor to convert 0 to 255. If you haven't seen the theme here yet; here it is. The MD layers should correct for platform-specific aberrations in the PCI implementation where possible. Adding compile-time options to MI code which indirectly relate exclusively to MD PCI issues is just the Wrong Thing to Do. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-arch" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200112140028.fBE0Sol04630>