Re: [mm-patch] bluetooth: use GFP_ATOMIC in *_sock_create's sk_alloc

[Marcel Holtmann <marcel@holtmann.org>]

Hi Frederik,

> > ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.18-rc2/2.6.18-rc2-mm1/
> > 
> 
> I think that the bluetooth-guard-bt_proto-with-rwlock.patch introduced the following
> BUG:
> [   43.232000] BUG: sleeping function called from invalid context at mm/slab.c:2903
> [   43.232000] in_atomic():1, irqs_disabled():0
> [   43.232000]  [<c0104114>] show_trace_log_lvl+0x197/0x1ba
> [   43.232000]  [<c010415e>] show_trace+0x27/0x29
> [   43.232000]  [<c010426e>] dump_stack+0x26/0x28
> [   43.232000]  [<c011ad1c>] __might_sleep+0xa2/0xaa
> [   43.232000]  [<c0173085>] __kmalloc+0x9c/0xb3
> [   43.232000]  [<c02f9295>] sk_alloc+0x1bc/0x1de
> [   43.232000]  [<c036d689>] hci_sock_create+0x42/0x8a
> [   43.236000]  [<c0366f40>] bt_sock_create+0xb5/0x154
> [   43.236000]  [<c02f69dc>] __sock_create+0x131/0x356
> [   43.236000]  [<c02f6c2f>] sock_create+0x2e/0x30
> [   43.236000]  [<c02f6c88>] sys_socket+0x27/0x53
> [   43.240000]  [<c02f7db5>] sys_socketcall+0xa9/0x277
> [   43.240000]  [<c0103131>] sysenter_past_esp+0x56/0x8d
> [   43.240000]  [<b7f38410>] 0xb7f38410

the comment from Max Krasnyansky (the original author) I got was this:

  As far as I remember there was some upper level locking that ensured
  that protected registration stuff. But it's been awhile so I may be
  completely wrong.

And Masatake YAMATO mentioned:

  It seems that lock_kernel/unlock_kernel was used in sys_init_module.
  However, now it is gone.

I haven't looked at the new module loading code to verify if we really
need to protect our socket registration or not.

Regards

Marcel