1f991bb365
I tested this on a WRT54GS and a WRT610N V1. The WRT610N does not boot when the kernel is too big, so I deactivated KALLSYMS Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
41 lines
1.2 KiB
Diff
41 lines
1.2 KiB
Diff
From: b.sander
|
|
Subject: [PATCH] pci: IDE fix
|
|
|
|
These are standard probing messages when using pdc202xx_old:
|
|
pdc202xx_old 0000:00:01.0: IDE controller (0x105a:0x0d30 rev 0x02)
|
|
PCI: Enabling device 0000:00:01.0 (0004 -> 0007)
|
|
PCI: Fixing up device 0000:00:01.0
|
|
0000:00:01.0: (U)DMA Burst Bit DISABLED Primary PCI Mode Secondary PCI Mode.
|
|
0000:00:01.0: FORCING BURST BIT 0x00->0x01 ACTIVE
|
|
pdc202xx_old 0000:00:01.0: 100% native mode on irq 6
|
|
|
|
With the default MAX_HWIFS value after above we get:
|
|
ide2: BM-DMA at 0x0400-0x0407
|
|
ide3: BM-DMA at 0x0408-0x040f
|
|
Probing IDE interface ide2...
|
|
hde: CF500, CFA DISK drive
|
|
|
|
As you can see it's ide2 + ide3 and hde.
|
|
|
|
With this patch applied we get:
|
|
ide0: BM-DMA at 0x0400-0x0407
|
|
ide1: BM-DMA at 0x0408-0x040f
|
|
Probing IDE interface ide0...
|
|
hda: CF500, CFA DISK drive
|
|
|
|
This fixes OpenWrt ticket #7061: https://dev.openwrt.org/ticket/7061
|
|
---
|
|
--- a/include/linux/ide.h
|
|
+++ b/include/linux/ide.h
|
|
@@ -204,7 +204,11 @@ static inline void ide_std_init_ports(st
|
|
hw->io_ports.ctl_addr = ctl_addr;
|
|
}
|
|
|
|
+#if defined CONFIG_BCM47XX
|
|
+# define MAX_HWIFS 2
|
|
+#else
|
|
#define MAX_HWIFS 10
|
|
+#endif
|
|
|
|
/*
|
|
* Now for the data we need to maintain per-drive: ide_drive_t
|