[PATCH] 2.5.44 and Secondary Channel Only configuration Kernel Ooops

[Jason Williams <jason_williams@suth.com>]

[-- Attachment #1: Type: text/plain, Size: 599 bytes --]

Here is a patch to correct the problem with the IDE code when you have
the Secondary Channel Only enabled in your BIOS.  If you have only the
secondary channel enabled, a null pointer get passed to the
ide_iomio_dma function as the part of the hwif var (more specifically
hwif->mate)  This patch puts in some extra sanity checks in that section
so make sure we don't hit the null.

Jason

P.S.  This is a continuation of a previous patch I submitted for 2.5.43
which wasn't committed and I didn't send that to the maintainer (Sorry
Andre) so I am resending it now as a patch to the 2.5.44 kernel.



[-- Attachment #2: Type: text/x-patch, Size: 1243 bytes --]

diff -u -r linux-2.5.44-vanilla/drivers/ide/ide-dma.c linux-2.5.44/drivers/ide/ide-dma.c
--- linux-2.5.44-vanilla/drivers/ide/ide-dma.c	2002-10-21 08:02:49.000000000 -0500
+++ linux-2.5.44/drivers/ide/ide-dma.c	2002-10-21 08:56:49.000000000 -0500
@@ -905,7 +905,12 @@
 		request_region(base+16, hwif->cds->extra, hwif->cds->name);
 		hwif->dma_extra = hwif->cds->extra;
 	}
-	hwif->dma_master = (hwif->channel) ? hwif->mate->dma_base : base;
+	if(!hwif->mate && hwif->channel){
+		hwif->dma_master=base;
+	}
+	else{
+		hwif->dma_master = (hwif->channel) ? hwif->mate->dma_base : base;
+	}
 	if (hwif->dma_base2) {
 		if (!check_mem_region(hwif->dma_base2, ports))
 			request_mem_region(hwif->dma_base2, ports, hwif->name);
@@ -925,8 +930,13 @@
 	if ((hwif->cds->extra) && (hwif->channel == 0)) {
 		request_region(base+16, hwif->cds->extra, hwif->cds->name);
 		hwif->dma_extra = hwif->cds->extra;
+	}	
+	if(!hwif->mate && hwif->channel){
+		hwif->dma_master=base;
+	}
+	else{
+		hwif->dma_master = (hwif->channel) ? hwif->mate->dma_base : base;
 	}
-	hwif->dma_master = (hwif->channel) ? hwif->mate->dma_base : base;
 	if (hwif->dma_base2) {
 		if (!request_region(hwif->dma_base2, ports, hwif->name))
 		{