Re: kernel BUG at /usr/src/linux-2.6.25/kernel/workqueue.c:224! (r8169?)

["Andrew Paprocki" <andrew@ishiboo.com>]

Francois,

On this kernel CONFIG_R8169_NAPI is disabled, so that code would not
even be compiled in. Is NAPI stable enough that I should be using it
now? (It is still marked experimental)

Thanks,
-Andrew

On Mon, Apr 21, 2008 at 4:34 PM, Francois Romieu <romieu@fr.zoreil.com> wrote:
> Andrew Paprocki <andrew@ishiboo.com> :
>
> > I just received a similar bug on workqueue.c:223 while I was not doing
>  > anything with the card. These bugs seem to have increased with
>  > frequency when running under 2.6.25. It used to take me some time to
>  > cause a problem to appear under 2.6.24, but under 2.6.25 they appear
>  > very readily. I was actually not touching the machine at all when this
>  > appeared on the console:
>
>  Does the patch below decrease the frequency of the bug ?
>
>  diff --git a/drivers/net/r8169.c b/drivers/net/r8169.c
>  index 3acfeea..6ee9db1 100644
>  --- a/drivers/net/r8169.c
>  +++ b/drivers/net/r8169.c
>  @@ -2398,8 +2398,6 @@ static void rtl8169_wait_for_quiescence(struct net_device *dev)
>         rtl8169_irq_mask_and_ack(ioaddr);
>
>   #ifdef CONFIG_R8169_NAPI
>  -       tp->intr_mask = 0xffff;
>  -       RTL_W16(IntrMask, tp->intr_event);
>         napi_enable(&tp->napi);
>   #endif
>   }
>
>  --
>  Ueimor
>
>