Date: Sat, 20 Jul 2002 13:49:17 -0700 From: Peter Wemm <peter@wemm.org> To: Scott Long <scott_long@btc.adaptec.com> Cc: Dag-Erling Smorgrav <des@ofug.org>, cvs-committers@freebsd.org, cvs-all@freebsd.org Subject: Re: cvs commit: src/sys/conf files.sparc64 Message-ID: <20020720204917.A0B23380A@overcee.wemm.org> In-Reply-To: <20020720202832.GA43959@hollin.btc.adaptec.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Scott Long wrote: > > Index: dev/aic7xxx/aic79xx_osm.h > > =================================================================== > > RCS file: /home/ncvs/src/sys/dev/aic7xxx/aic79xx_osm.h,v > > retrieving revision 1.2 > > diff -u -r1.2 aic79xx_osm.h > > --- dev/aic7xxx/aic79xx_osm.h 6 Jun 2002 16:35:57 -0000 1.2 > > +++ dev/aic7xxx/aic79xx_osm.h 20 Jul 2002 17:51:55 -0000 > > @@ -38,9 +38,6 @@ > > #define _AIC79XX_FREEBSD_H_ > > > > #include <opt_aic79xx.h> /* for config options */ > > -#ifndef NPCI > > -#include <pci.h> > > -#endif > > > > #include <sys/param.h> > > #include <sys/systm.h> > > @@ -60,10 +57,8 @@ > > > > #include <sys/rman.h> > > > > -#if NPCI > 0 > > #include <pci/pcireg.h> > > #include <pci/pcivar.h> > > -#endif > > > > #include <cam/cam.h> > > #include <cam/cam_ccb.h> > > While this is probably harmless, there might be non-pci attachments in the > future that break this. Heh, right now it is a NOP for this device.. And see below. > > Index: dev/aic7xxx/aic7xxx_osm.h > > =================================================================== > > RCS file: /home/ncvs/src/sys/dev/aic7xxx/aic7xxx_osm.h,v > > retrieving revision 1.14 > > diff -u -r1.14 aic7xxx_osm.h > > --- dev/aic7xxx/aic7xxx_osm.h 26 Apr 2002 22:48:20 -0000 1.14 > > +++ dev/aic7xxx/aic7xxx_osm.h 20 Jul 2002 17:52:06 -0000 > > @@ -38,9 +38,6 @@ > > #define _AIC7XXX_FREEBSD_H_ > > > > #include <opt_aic7xxx.h> /* for config options */ > > -#ifndef NPCI > > -#include <pci.h> > > -#endif > > > > #include <sys/param.h> > > #include <sys/systm.h> > > @@ -51,12 +48,10 @@ > > #include <sys/malloc.h> > > #include <sys/queue.h> > > > > -#if NPCI > 0 > > #define AHC_PCI_CONFIG 1 > > #ifdef AHC_ALLOW_MEMIO > > #include <machine/bus_memio.h> > > #endif > > -#endif > > #include <machine/bus_pio.h> > > #include <machine/bus.h> > > #include <machine/endian.h> > > @@ -65,10 +60,8 @@ > > > > #include <sys/rman.h> > > > > -#if NPCI > 0 > > #include <pci/pcireg.h> > > #include <pci/pcivar.h> > > -#endif > > > > #include <cam/cam.h> > > #include <cam/cam_ccb.h> > > This will break the 274x and 284x cards (EISA and VLB, respectively). No, it wont, becase there is a seperate option: #ifdef AHC_ALLOW_MEMIO #include <machine/bus_memio.h> #endif And even if AHC_ALLOW_MEMIO is specified in the kernel config, whether it uses memio or pio is a property of the bus space handle/tag. What having both of the methods compiled in does is cause a runtime switch between one or the other. 274x and 284x (I have a 2742T system still) are (at face value) unaffected by this change. Besides, my 2742T system has PCI as well, so NPCI *is* 1 on it, and it Works Just Fine. In fact, GENERIC has NPCI = 1 as well so this code is used in 99% of freebsd installs "out there". > Is getting rid of NPCI an urgent issue? The aic7xxx can probably be fixed, > but not without some effort. I need to resurrect the work I did on > separating the driver into descrete attachment front-ends. This doesn't break it. Cheers, -Peter -- Peter Wemm - peter@wemm.org; peter@FreeBSD.org; peter@yahoo-inc.com "All of this is for nothing if we don't go to the stars" - JMS/B5 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe cvs-all" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20020720204917.A0B23380A>