Re: Serverworks Oops with the very latest git kernel...

From: Jeff Garzik
Date: Thu Jan 10 2008 - 21:24:45 EST


Trond Myklebust wrote:
Hi,

I'm getting the following Oops on boot with kernel 2.6.24-rc7-g88fb61e4.

Starting udev: BUG: unable to handle kernel paging request at virtual address 2d6d3a65
printing eip: c04e1af1 *pde = 00000000 Oops: 0000 [#1] PREEMPT SMP Modules linked in: evdev evbug rtc_cmos pcspkr floppy rtc_core rtc_lib shpchp pci_hotplug serverworks tg3 generic i2c_piix4 i2c_core

Pid: 1820, comm: modprobe Not tainted (2.6.24-rc7-g88fb61e4 #2)
EIP: 0060:[<c04e1af1>] EFLAGS: 00210286 CPU: 0
EIP is at strstr+0x11/0x34
EAX: 00000000 EBX: 2d6d3a65 ECX: ffffffff EDX: 2d6d3a65
ESI: c0513f75 EDI: 2d6d3a65 EBP: f65bccec ESP: f65bcce0
DS: 007b ES: 007b FS: 00d8 GS: 0033 SS: 0068
Process modprobe (pid: 1820, ti=f65bc000 task=f795a2f0 task.ti=f65bc000)
Stack: f8921a44 00000001 00000000 f65bcd00 c054a0b9 f892e394 f892e080 f892e5f8 f65bcd20 f892133b c04d7c9d f892da80 f891d339 f892e394 f892e080 f892e5f8 f65bcd64 f891d11f f891d339 f891bc38 f892e080 00000000 00000001 00000000 Call Trace:
[<c0405170>] show_trace_log_lvl+0x1a/0x2f
[<c0405222>] show_stack_log_lvl+0x9d/0xa5
[<c04052cd>] show_registers+0xa3/0x1df
[<c040552a>] die+0x121/0x204
[<c05b40dd>] do_page_fault+0x557/0x63e
[<c05b278a>] error_code+0x72/0x78
[<c054a0b9>] dmi_check_system+0x2f/0x58
[<f892133b>] ide_acpi_init+0x28/0x195 [ide_core]
[<f891d11f>] hwif_init+0x324/0x37c [ide_core]
[<f891d29a>] ide_device_add+0x3b/0xa6 [ide_core]
[<f891ee9a>] ide_setup_pci_device+0x36/0x40 [ide_core]
[<f885f11c>] svwks_init_one+0x5b/0x62 [serverworks]
[<c04eada1>] pci_device_probe+0x39/0x5b
[<c05369a8>] driver_probe_device+0xe9/0x16a
[<c0536b43>] __driver_attach+0x6c/0xa5
[<c0535e92>] bus_for_each_dev+0x36/0x5b
[<c05367f3>] driver_attach+0x19/0x1b
[<c05361b1>] bus_add_driver+0x73/0x1aa
[<c0536d29>] driver_register+0x67/0x6c
[<c04eaef9>] __pci_register_driver+0x56/0x83
[<f890f017>] svwks_ide_init+0x17/0x19 [serverworks]
[<c0447f1e>] sys_init_module+0x15da/0x1715
[<c04040b2>] sysenter_past_esp+0x5f/0xa5
=======================
Code: e5 89 c1 89 c8 eb 06 80 38 00 74 07 40 4a 83 fa ff 75 f4 29 c8 5d c3 90 90 90 55 89 e5 57 56 53 89 c6 89 d3 31 c0 83 c9 ff 89 EIP: [<c04e1af1>] strstr+0x11/0x34 SS:ESP 0068:f65bcce0
---[ end trace f0316f13333e9ff3 ]---
udevd-event[1819]: run_program: '/sbin/modprobe' abnormal exit


Any ideas? As far as I can see, 2.6.24-rc7 itself is good, so I suspect
the recent merge by Bart.

I've set up a bugzilla entry for the regression on
http://bugzilla.kernel.org/show_bug.cgi?id=9728

Does this patch fix it?

Looks like an unterminated DMI table...

Jeff


diff --git a/drivers/ide/ide-acpi.c b/drivers/ide/ide-acpi.c
index fe6768a..8b23ea5 100644
--- a/drivers/ide/ide-acpi.c
+++ b/drivers/ide/ide-acpi.c
@@ -85,6 +85,8 @@ static const struct dmi_system_id ide_acpi_dmi_table[] = {
DMI_MATCH(DMI_BIOS_VERSION, "KAM1.60")
},
},
+
+ { }
};

static int ide_acpi_blacklist(void)