2009-04-09 02:19:23

by Justin P. Mattock

[permalink] [raw]
Subject: BUG kmalloc-2048: Invalid object pointer 0xf656ad80

doing a:
sudo /sbin/rmmod wl
(generated a bug message)

[10018.950114] =============================================================================
[10018.950118] BUG kmalloc-2048: Invalid object pointer 0xf656ad80
[10018.950120] -----------------------------------------------------------------------------
[10018.950121]
[10018.950124] INFO: Slab 0xc16cad00 objects=15 used=14 fp=0xf656a8f0
flags=0x40002083
[10018.950128] Pid: 20025, comm: rmmod Tainted: P 2.6.29 #10
[10018.950131] Call Trace:
[10018.950139] [<c017db89>] slab_err+0x51/0x58
[10018.950143] [<c017e812>] ? __slab_free+0x26f/0x2a7
[10018.950147] [<c017dbaf>] ? slab_pad_check+0x1f/0xbd
[10018.950152] [<c017e83d>] __slab_free+0x29a/0x2a7
[10018.950186] [<f93f8a93>] ? osl_mfree+0x11/0x13 [wl]
[10018.950190] [<c017f59e>] kfree+0x93/0xa9
[10018.950220] [<f93f8a93>] ? osl_mfree+0x11/0x13 [wl]
[10018.950249] [<f93f8a93>] osl_mfree+0x11/0x13 [wl]
[10018.950279] [<f93f5cca>] wl_free_if+0x65/0x77 [wl]
[10018.950309] [<f93f5d0f>] wl_free+0x33/0x17f [wl]
[10018.950313] [<c017f701>] ? kmem_cache_free+0x71/0x87
[10018.950318] [<c01b9006>] ? release_sysfs_dirent+0x6c/0x81
[10018.950322] [<c01b9006>] ? release_sysfs_dirent+0x6c/0x81
[10018.950326] [<c01b9200>] ? sysfs_addrm_finish+0x1bf/0x1d2
[10018.950331] [<c022faac>] ? _raw_spin_lock+0x53/0xfb
[10018.950337] [<c03cc23f>] ? _spin_lock_bh+0x16/0x18
[10018.950341] [<c022fa55>] ? _raw_spin_unlock+0x74/0x78
[10018.950370] [<f93f8ca9>] wl_remove+0x42/0x58 [wl]
[10018.950376] [<c023a208>] pci_device_remove+0x19/0x39
[10018.950381] [<c029b427>] __device_release_driver+0x50/0x6d
[10018.950386] [<c029b4b4>] driver_detach+0x70/0x94
[10018.950390] [<c029aa26>] bus_remove_driver+0x63/0x7f
[10018.950394] [<c029b847>] driver_unregister+0x28/0x2c
[10018.950399] [<c023a3fe>] pci_unregister_driver+0x21/0x62
[10018.950427] [<f93f8c65>] wl_module_exit+0xd/0xf [wl]
[10018.950433] [<c01469eb>] sys_delete_module+0x18e/0x1e7
[10018.950438] [<c017219d>] ? remove_vma+0x41/0x47
[10018.950442] [<c0172d24>] ? do_munmap+0x1f1/0x20b
[10018.950446] [<c015397e>] ? audit_syscall_entry+0x15e/0x180
[10018.950453] [<c010301b>] sysenter_do_call+0x12/0x2f
[10018.950458] [<c03c0000>] ? workqueue_cpu_callback+0xc0/0xd6
[10018.950461] FIX kmalloc-2048: Object at 0xf656ad80 not freed
[10018.950745] wl 0000:03:00.0: PCI INT A disabled


The kernel is tainted with nvidia, and a broadcom wl.

--
Justin P. Mattock


2009-04-09 02:38:23

by Bryan Donlan

[permalink] [raw]
Subject: Re: BUG kmalloc-2048: Invalid object pointer 0xf656ad80

On Wed, Apr 8, 2009 at 10:19 PM, Justin Mattock <[email protected]> wrote:
> doing a:
> sudo /sbin/rmmod wl
> (generated a bug message)
[snip]
> The kernel is tainted with nvidia, and a broadcom wl.

Would this broadcom wl be the very module you're trying to remove?
What makes you think this isn't a bug in this wl binary module? From
the looks of the backtrace it's dying due to a free of a wild pointer
right in the middle of the wl module - it certainly doesn't look like
it came from some other subsystem...

2009-04-09 02:51:15

by Justin P. Mattock

[permalink] [raw]
Subject: Re: BUG kmalloc-2048: Invalid object pointer 0xf656ad80

On Wed, Apr 8, 2009 at 7:38 PM, Bryan Donlan <[email protected]> wrote:
> On Wed, Apr 8, 2009 at 10:19 PM, Justin Mattock <[email protected]> wrote:
>> doing a:
>> sudo /sbin/rmmod wl
>> (generated a bug message)
> [snip]
>> The kernel is tainted with nvidia, and a broadcom wl.
>
> Would this broadcom wl be the very module you're trying to remove?
> What makes you think this isn't a bug in this wl binary module? From
> the looks of the backtrace it's dying due to a free of a wild pointer
> right in the middle of the wl module - it certainly doesn't look like
> it came from some other subsystem...
>

yup,
remove wl, in exchange
for ndiswrapper. and poof
this showed up..

regards,

--
Justin P. Mattock

2009-04-09 03:00:44

by Bryan Donlan

[permalink] [raw]
Subject: Re: BUG kmalloc-2048: Invalid object pointer 0xf656ad80

On Wed, Apr 8, 2009 at 10:50 PM, Justin Mattock <[email protected]> wrote:
> On Wed, Apr 8, 2009 at 7:38 PM, Bryan Donlan <[email protected]> wrote:
>> On Wed, Apr 8, 2009 at 10:19 PM, Justin Mattock <[email protected]> wrote:
>>> doing a:
>>> sudo /sbin/rmmod wl
>>> (generated a bug message)
>> [snip]
>>> The kernel is tainted with nvidia, and a broadcom wl.
>>
>> Would this broadcom wl be the very module you're trying to remove?
>> What makes you think this isn't a bug in this wl binary module? From
>> the looks of the backtrace it's dying due to a free of a wild pointer
>> right in the middle of the wl module - it certainly doesn't look like
>> it came from some other subsystem...
>>
>
> yup,
> remove wl, in exchange
> for ndiswrapper. and poof
> this showed up..

The bug came up in the wl module's removal code - I'd suggest just not
loading it in the first place.

2009-04-09 03:07:05

by Justin P. Mattock

[permalink] [raw]
Subject: Re: BUG kmalloc-2048: Invalid object pointer 0xf656ad80

On Wed, Apr 8, 2009 at 8:00 PM, Bryan Donlan <[email protected]> wrote:
> On Wed, Apr 8, 2009 at 10:50 PM, Justin Mattock <[email protected]> wrote:
>> On Wed, Apr 8, 2009 at 7:38 PM, Bryan Donlan <[email protected]> wrote:
>>> On Wed, Apr 8, 2009 at 10:19 PM, Justin Mattock <[email protected]> wrote:
>>>> doing a:
>>>> sudo /sbin/rmmod wl
>>>> (generated a bug message)
>>> [snip]
>>>> The kernel is tainted with nvidia, and a broadcom wl.
>>>
>>> Would this broadcom wl be the very module you're trying to remove?
>>> What makes you think this isn't a bug in this wl binary module? From
>>> the looks of the backtrace it's dying due to a free of a wild pointer
>>> right in the middle of the wl module - it certainly doesn't look like
>>> it came from some other subsystem...
>>>
>>
>> yup,
>> remove wl, in exchange
>> for ndiswrapper. and poof
>> this showed up..
>
> The bug came up in the wl module's removal code - I'd suggest just not
> loading it in the first place.
>

The module sounded like a good idea,
the problem I have(not sure if this is kernel related)
is after the network is up, loading anything takes
a good 30's, then after about 15min later my router will
do a reset(disconnecting any machine connected). could be something missing
with wpasupplicant, kernel, some lib, who knows...
(at this point my main goal is to get the rest of
the system built before dealing with quirks).

regards,

--
Justin P. Mattock