PATCH: (Discussion) Stop IDE legacy ISA probes on PCI systems

PATCH: (Discussion) Stop IDE legacy ISA probes on PCI systems

[Alan Cox]

We don't generally want to poke around legacy ISA ide2..ide5 on a system
with PCI , it tends to cause long delays. 

diff -u --new-file --recursive --exclude-from /usr/src/exclude linux.vanilla-2.6.10/include/asm-i386/ide.h linux-2.6.10/include/asm-i386/ide.h
--- linux.vanilla-2.6.10/include/asm-i386/ide.h	2004-12-25 21:13:51.000000000 +0000
+++ linux-2.6.10/include/asm-i386/ide.h	2004-12-26 17:08:27.541915144 +0000
@@ -44,10 +44,14 @@
 	switch (index) {
 		case 0:	return 0x1f0;
 		case 1:	return 0x170;
-		case 2: return 0x1e8;
-		case 3: return 0x168;
-		case 4: return 0x1e0;
-		case 5: return 0x160;
+		
+		if(pci_find_device(PCI_ANY_ID, PCI_ANY_ID, NULL) == NULL) {
+			case 2: return 0x1e8;
+			case 3: return 0x168;
+			case 4: return 0x1e0;
+			case 5: return 0x160;
+		}
+
 		default:
 			return 0;
 	}

Re: PATCH: (Discussion) Stop IDE legacy ISA probes on PCI systems

[Adam Sampson]

Alan Cox <alan@lxorguk.ukuu.org.uk> writes:

>  	switch (index) {
>  		case 0:	return 0x1f0;
>  		case 1:	return 0x170;
> +		
> +		if(pci_find_device(PCI_ANY_ID, PCI_ANY_ID, NULL) == NULL) {
> +			case 2: return 0x1e8;
> +			case 3: return 0x168;
> +			case 4: return 0x1e0;
> +			case 5: return 0x160;
> +		}
> +
>  		default:
>  			return 0;
>  	}

I don't think that code will have the intended effect, unless your
GCC has some funny ideas about switch statements...

-- 
Adam Sampson <azz@us-lot.org>                        <http://offog.org/>

Re: PATCH: (Discussion) Stop IDE legacy ISA probes on PCI systems

[Valdis.Kletnieks]

[-- Attachment #1: Type: text/plain, Size: 869 bytes --]

On Mon, 27 Dec 2004 21:15:26 GMT, Adam Sampson said:
> Alan Cox <alan@lxorguk.ukuu.org.uk> writes:
> 
> >  	switch (index) {
> >  		case 0:	return 0x1f0;
> >  		case 1:	return 0x170;
> > +		
> > +		if(pci_find_device(PCI_ANY_ID, PCI_ANY_ID, NULL) == NULL) {
> > +			case 2: return 0x1e8;
> > +			case 3: return 0x168;
> > +			case 4: return 0x1e0;
> > +			case 5: return 0x160;
> > +		}
> > +
> >  		default:
> >  			return 0;
> >  	}
> 
> I don't think that code will have the intended effect, unless your
> GCC has some funny ideas about switch statements...

Those who don't understand the Duff Device are doomed to re-invent it, poorly. ;)

(Yes Alan, I know *you* know C well enough, but I just looked at that block
of code and it looked *so* much like Tom Duff's that I had to say it. ;)

For those not in the know:

http://www.lysator.liu.se/c/duffs-device.html

[-- Attachment #2: Type: application/pgp-signature, Size: 226 bytes --]

Re: PATCH: (Discussion) Stop IDE legacy ISA probes on PCI systems

[Linus Torvalds]

On Mon, 27 Dec 2004, Adam Sampson wrote:
> 
> I don't think that code will have the intended effect, unless your
> GCC has some funny ideas about switch statements...

Indeed. That if-statement is unreachable and has no effect.

		Linus

Re: PATCH: (Discussion) Stop IDE legacy ISA probes on PCI systems

[Coywolf Qi Hunt]

This should solve the problem. 

              coywolf


--- linux-2.6.10/include/asm-i386/ide.h 2004-10-19 05:53:13.000000000 +0800
+++ linux-2.6.10-cy/include/asm-i386/ide.h      2004-12-28
11:45:38.000000000 +0800
@@ -44,13 +44,18 @@ static __inline__ unsigned long ide_defa
        switch (index) {
                case 0: return 0x1f0;
                case 1: return 0x170;
-               case 2: return 0x1e8;
-               case 3: return 0x168;
-               case 4: return 0x1e0;
-               case 5: return 0x160;
-               default:
-                       return 0;
        }
+
+       if (pci_find_device(PCI_ANY_ID, PCI_ANY_ID, NULL) == NULL) {
+               switch (index) {
+                       case 2: return 0x1e8;
+                       case 3: return 0x168;
+                       case 4: return 0x1e0;
+                       case 5: return 0x160;
+               }
+       }
+
+       return 0;
 }

 #define IDE_ARCH_OBSOLETE_INIT


On Mon, 27 Dec 2004 16:29:52 -0800 (PST), Linus Torvalds
<torvalds@osdl.org> wrote:
> 
> 
> On Mon, 27 Dec 2004, Adam Sampson wrote:
> >
> > I don't think that code will have the intended effect, unless your
> > GCC has some funny ideas about switch statements...
> 
> Indeed. That if-statement is unreachable and has no effect.
> 
>                 Linus
> -

-- 
Coywolf Qi Hunt
Homepage http://sosdg.org/~coywolf/