2.5.53 AGP module oops

2.5.53 AGP module oops

Post by Mikael Pettersso » Sun, 29 Dec 2002 02:50:07



Dave,

Kernel 2.5.53 changed the AGP config options so that
CONFIG_AGP=m forces =m on the HW-specific config. When I
insmod the HW-specific module (intel-agp in my case)
after insmod agpgart, the kernel immediately oopses:

Unable to handle kernel NULL pointer dereference at virtual address 00000074
 printing eip:
d089800c
*pde = 00000000
Oops: 0000
CPU:    0
EIP:    0060:[<d089800c>]    Not tainted
EFLAGS: 00010206
EIP is at 0xd089800c
eax: c0221714   ebx: cff4d000   ecx: ce606080   edx: c0221750
esi: 00000000   edi: cff4d04c   ebp: 00000000   esp: ce367ef0
ds: 0068   es: 0068   ss: 0068
Process insmod (pid: 577, threadinfo=ce366000 task=ce606080)
Stack: d08a60ba cff4d000 00000000 cff4d04c cff4d000 00000000 d08a62ac cff4d000
       cff4d000 d089ad10 cff4d000 d08addc8 c016af49 cff4d000 d089ae50 cff4d04c
       d08addc8 ffffffed d08addc8 d08adda0 c0171677 cff4d04c cff4d04c cff4d054
Call Trace:
 [<d08a60ba>] agp_backend_initialize+0x16/0x168 [agpgart]
 [<d08a62ac>] agp_register_driver+0x28/0xa0 [agpgart]
 [<d089ad10>] 0xd089ad10
 [<d08addc8>] agp_intel_pci_driver+0x28/0x9c [intel_agp]
 [<c016af49>] pci_device_probe+0x41/0x5c
 [<d089ae50>] 0xd089ae50
 [<d08addc8>] agp_intel_pci_driver+0x28/0x9c [intel_agp]
 [<d08addc8>] agp_intel_pci_driver+0x28/0x9c [intel_agp]
 [<d08adda0>] agp_intel_pci_driver+0x0/0x9c [intel_agp]
 [<c0171677>] bus_match+0x37/0x68
 [<c0171743>] driver_attach+0x37/0x60
 [<d08addc8>] agp_intel_pci_driver+0x28/0x9c [intel_agp]
 [<d08adb62>] __module_pci_device_size+0x12/0x30 [intel_agp]
 [<d08addc8>] agp_intel_pci_driver+0x28/0x9c [intel_agp]
 [<c01719de>] bus_add_driver+0x92/0xb4
 [<d08addc8>] agp_intel_pci_driver+0x28/0x9c [intel_agp]
 [<d08adde8>] agp_intel_pci_driver+0x48/0x9c [intel_agp]
 [<c0171d9c>] driver_register+0x34/0x38
 [<d08addc8>] agp_intel_pci_driver+0x28/0x9c [intel_agp]
 [<c016b042>] pci_register_driver+0x42/0x54
 [<d08addc8>] agp_intel_pci_driver+0x28/0x9c [intel_agp]
 [<d089ad3e>] 0xd089ad3e
 [<d08adda0>] agp_intel_pci_driver+0x0/0x9c [intel_agp]
 [<c01247e9>] sys_init_module+0xe9/0x170
 [<c0108897>] syscall_call+0x7/0xb

Code: 69 6e 74 65 6c 5f 61 67 70 00 00 00 00 00 00 00 00 00 00 00

The stack trace shows that we should be in agp_find_max() as
called from agp_backend_initialize(). However, agp_find_max()
is __init and its code has already been removed at this point
(since agpgart and intel-agp are separate modules), causing
the kernel to execute random code and eventually oops.

The patch below works around this by changing agpgart's __init
code & data to normal code & data. Tested, works for me.

/Mikael

--- linux-2.5.53/drivers/char/agp/backend.c.~1~ 2002-12-24 13:53:50.000000000 +0100

        int agp;
 };

-static struct agp_max_table maxes_table[9] __initdata =
+static struct agp_max_table maxes_table[9] =
 {
        {0, 0},

        {4096, 3932}
 };

-static int __init agp_find_max (void)
+static int agp_find_max (void)
 {
        long memory, index, result;

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in

More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

 
 
 

2.5.53 AGP module oops

Post by Dave Jone » Sun, 29 Dec 2002 21:00:11


 > The stack trace shows that we should be in agp_find_max() as
 > called from agp_backend_initialize(). However, agp_find_max()
 > is __init and its code has already been removed at this point
 > (since agpgart and intel-agp are separate modules), causing
 > the kernel to execute random code and eventually oops.
 >
 > The patch below works around this by changing agpgart's __init
 > code & data to normal code & data. Tested, works for me.

Looks good to me, applied.

Thanks.

                Dave

--
| Dave Jones.        http://www.codemonkey.org.uk
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in

More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/