2.4.17-pre2 & PCMCIA Errors

2.4.17-pre2 & PCMCIA Errors

[Alan Ford]

Just tried 2.4.17-pre2 (was previously on 2.4.16-pre1) and when pcmcia-cs is
started on bootup, the following happens:

cs: IO port probe 0x0c00-0x0cff: clean.
cs: IO port probe 0x0100-0x04ff: excluding 0x4d0-0x4d7
cs: IO port probe 0x0a00-0x0aff: clean.
cs: memory probe 0xa0000000-0xa0ffffff:<1>Unable to handle kernel NULL pointer
dereference at virtual address 00000004
 printing eip:
c0118781
*pde = 00000000
Oops: 0000
CPU:    0
EIP:    0010:[<c0118781>]    Not tainted
EFLAGS: 00010206
eax: 00000000   ebx: cf57d988   ecx: 00000000   edx: a00fffff
esi: a00fffff   edi: a0000000   ebp: 00100000   esp: cf57d968
ds: 0018   es: 0018   ss: 0018
Process cardmgr (pid: 238, stackpage=cf57d000)
Stack: cf57d988 a0000000 a0000000 c0118854 00000000 cf57d988 cfdc1bc0 a0000000
       00000200 a0000000 a00fffff cfdc1bc0 c01bd035 a0000000 00100000 c01bd03e
       00000000 a0000000 00100000 cfdc1bc0 00000000 cff4e000 c01badb4 a1000000
Call Trace: [<c0118854>] [<c01bd035>] [<c01bd03e>] [<c01badb4>] [<c01bd1cb>]
   [<c01bac34>] [<c01badb4>] [<c01badb4>] [<c01bd180>] [<c01bac34>] [<c01badb4>]
   [<c01badb4>] [<c01bd180>] [<c01bac34>] [<c01badb4>] [<c01bd225>] [<c01bac34>]
   [<c01badb4>] [<c01bae22>] [<c01bac34>] [<c01badb4>] [<c01ba9bc>] [<c01bb012>]
   [<c01bb4eb>] [<c01bb30a>] [<c01bc9a5>] [<c018167b>] [<c01c96b7>] [<c018218e>]
   [<c0185eac>] [<c0185496>] [<c0114646>] [<c01146ab>] [<c0114781>] [<c01149a3>]
   [<c01148f4>] [<c01bcf8b>] [<c0140002>] [<c01bf44f>] [<c019d5eb>] [<c01a3b10>]
   [<c0107cff>] [<c019be15>] [<c015cab5>] [<c015cbc3>] [<c015e34c>] [<c015e575>]
   [<c015ea0d>] [<c01303c3>] [<c0130484>] [<c0130694>] [<c013089f>] [<c0130260>]
   [<c0129838>] [<c0129bdd>] [<c012014a>] [<c01bf08c>] [<c013a828>] [<c013a862>]
   [<c013acbf>] [<c01bef57>] [<c0139e77>] [<c0106b0b>]

Code: 3b 79 04 72 05 3b 71 08 76 05 89 c8 eb 29 90 8d 51 18 8b 02


If I run cardctl, I'll get the same error ("Unable to handle kernel NULL
pointer dereference at virtual address 00000004")

Can anybody enlighten me? Is this a kernel bug or have I got something
misconfigured that just happened to work before (if so, can that "feature"
be put back please? :) )

Am using Debian with pcmcia-cs 3.1.29-3. Laptop is a Dell Insprion 8K 
(CardBus controller is a TI PCI4451).

-- 
Alan Ford * alan@whirlnet.co.uk 

Re: 2.4.17-pre2 & PCMCIA Errors

[Paul Mackerras]

Alan Ford writes:

> Just tried 2.4.17-pre2 (was previously on 2.4.16-pre1) and when pcmcia-cs is
> started on bootup, the following happens:
> 
> cs: IO port probe 0x0c00-0x0cff: clean.
> cs: IO port probe 0x0100-0x04ff: excluding 0x4d0-0x4d7
> cs: IO port probe 0x0a00-0x0aff: clean.
> cs: memory probe 0xa0000000-0xa0ffffff:<1>Unable to handle kernel NULL pointer
> dereference at virtual address 00000004

Please try this patch and let me know whether it fixes the problem.

Paul.

diff -urN linux-2.4.17-pre2/drivers/pcmcia/rsrc_mgr.c pmac/drivers/pcmcia/rsrc_mgr.c
--- linux-2.4.17-pre2/drivers/pcmcia/rsrc_mgr.c	Sat Dec  1 15:49:24 2001
+++ pmac/drivers/pcmcia/rsrc_mgr.c	Mon Dec  3 14:28:16 2001
@@ -107,17 +107,19 @@
 static struct resource *resource_parent(unsigned long b, unsigned long n,
 					int flags, struct pci_dev *dev)
 {
-	struct resource res;
+	struct resource res, *pr;
 
-	if (dev == NULL) {
-		if (flags & IORESOURCE_MEM)
-			return &iomem_resource;
-		return &ioport_resource;
+	if (dev != NULL) {
+		res.start = b;
+		res.end = b + n - 1;
+		res.flags = flags;
+		pr = pci_find_parent_resource(dev, &res);
+		if (pr)
+			return pr;
 	}
-	res.start = b;
-	res.end = b + n - 1;
-	res.flags = flags;
-	return pci_find_parent_resource(dev, &res);
+	if (flags & IORESOURCE_MEM)
+		return &iomem_resource;
+	return &ioport_resource;
 }
 
 static inline int check_io_resource(unsigned long b, unsigned long n,

Re: 2.4.17-pre2 & PCMCIA Errors

[Alan Ford]

On Mon, Dec 03, 2001 at 03:29:57PM +1100, Paul Mackerras wrote:
> Alan Ford writes:
> 
> > Just tried 2.4.17-pre2 (was previously on 2.4.16-pre1) and when pcmcia-cs is
> > started on bootup, the following happens:
> > 
> > cs: IO port probe 0x0c00-0x0cff: clean.
> > cs: IO port probe 0x0100-0x04ff: excluding 0x4d0-0x4d7
> > cs: IO port probe 0x0a00-0x0aff: clean.
> > cs: memory probe 0xa0000000-0xa0ffffff:<1>Unable to handle kernel NULL pointer
> > dereference at virtual address 00000004
> 
> Please try this patch and let me know whether it fixes the problem.

Thank you, that appears to fix the problem.

The disks are also unmounted correctly now, so that problem was probably
caused by the kernel getting in a mess (excuse my non-technical language)
over this pcmcia stuff.

-- 
Alan Ford * alan@whirlnet.co.uk