2002-11-16 09:47:30

by Zwane Mwaikambo

Subject: [PATCH][2.5] USB core/urb.c triggers slab bugcheck

This patch is also required to get my box to boot

kernel BUG at mm/slab.c:1619!
invalid operand: 0000

CPU: 0
EIP: 0060:[<c0147a2d>] Not tainted
EFLAGS: 00010283
EIP is at cache_alloc_debugcheck_after+0x4d/0xa0
eax: 04000409 ebx: c15bd398 ecx: 00000068 edx: c15bd400
esi: c7fff320 edi: 00000246 ebp: 000001d0 esp: c148fe50
ds: 0068 es: 0068 ss: 0068
Process khubd (pid: 5, threadinfo=c148e000 task=c7f75940)
Stack: c7fff320 c148e000 c0146b82 c7fff320 000001d0 c15bd398 c15a7c00 c7e0f244
80000200 00000000 c031bd66 00000060 000001d0 c15a7c00 c031c29d 00000000
000001d0 00000028 c031c38f c15a7c00 80000200 c7e0f244 00000000 00000000
Call Trace:
[<c0146b82>] kmalloc+0x82/0xb0
[<c031bd66>] usb_alloc_urb+0x16/0x60
[<c031c29d>] usb_internal_control_msg+0xd/0x80
[<c031c38f>] usb_control_msg+0x7f/0xa0
[<c031d04f>] usb_set_configuration+0x7f/0xe0
[<c031793e>] usb_new_device+0x1be/0x3f0
[<c03199df>] usb_hub_port_connect_change+0x17f/0x280
[<c0319df0>] usb_hub_events+0x310/0x3d0
[<c0319ee5>] usb_hub_thread+0x35/0xe0
[<c011f7c0>] default_wake_function+0x0/0x40
[<c0319eb0>] usb_hub_thread+0x0/0xe0
[<c0108e15>] kernel_thread_helper+0x5/0x10

Code: 0f 0b 53 06 41 8f 45 c0 83 c3 04 8b 46 68 85 c0 74 06 f6 46

Index: linux-2.5.47-zm1/drivers/usb/core/urb.c
===================================================================
RCS file: /build/cvsroot/linux-2.5.47/drivers/usb/core/urb.c,v
retrieving revision 1.1.1.1
diff -u -r1.1.1.1 urb.c
--- linux-2.5.47-zm1/drivers/usb/core/urb.c 11 Nov 2002 03:57:41 -0000 1.1.1.1
+++ linux-2.5.47-zm1/drivers/usb/core/urb.c 16 Nov 2002 09:32:47 -0000
@@ -33,17 +33,15 @@
{
struct urb *urb;

- urb = (struct urb *)kmalloc(sizeof(struct urb) +
- iso_packets * sizeof(struct usb_iso_packet_descriptor),
- mem_flags);
- if (!urb) {
+ urb = kmalloc(sizeof(*urb)
+ + (iso_packets * sizeof(struct usb_iso_packet_descriptor)),
+ mem_flags);
+ if (urb) {
+ memset(urb, 0, sizeof(*urb));
+ atomic_set(&urb->count, 1);
+ spin_lock_init(&urb->lock);
+ } else
err("alloc_urb: kmalloc failed");
- return NULL;
- }
-
- memset(urb, 0, sizeof(*urb));
- urb->count = (atomic_t)ATOMIC_INIT(1);
- spin_lock_init(&urb->lock);

return urb;
}
--
function.linuxpower.ca

2002-11-16 10:14:18

by Oliver Neukum

Subject: Re: [PATCH][2.5] USB core/urb.c triggers slab bugcheck

Am Samstag, 16. November 2002 10:57 schrieb Zwane Mwaikambo:
> This patch is also required to get my box to boot

Why is this needed ? I might be stupid, but it seems as if this patch
changes nothing.

Regards
Oliver

2002-11-16 10:27:37

by Zwane Mwaikambo

Subject: Re: [PATCH][2.5] USB core/urb.c triggers slab bugcheck

On Sat, 16 Nov 2002, Zwane Mwaikambo wrote:

> This patch is also required to get my box to boot
>
> kernel BUG at mm/slab.c:1619!
> invalid operand: 0000

Sorry please ignore this one, the first patch fixed the boot problem, i
hadn't tried booting with it alone. This must have been fallout.

Zwane

--
function.linuxpower.ca

2002-11-16 10:35:59

by Zwane Mwaikambo

Subject: Re: [PATCH][2.5] USB core/urb.c triggers slab bugcheck

On Sat, 16 Nov 2002, Oliver Neukum wrote:

> Am Samstag, 16. November 2002 10:57 schrieb Zwane Mwaikambo:
> > This patch is also required to get my box to boot
>
> Why is this needed ? I might be stupid, but it seems as if this patch
> changes nothing.

Au contraire, i'm the idiot, i had debugged all the oopses in one sitting
without booting each 'fix', i was pretty certain that bug had been triggered
seperately in an earlier boot.

Zwane
--
function.linuxpower.ca