netdev
[Top] [All Lists]

Re: [PATCH 6/28] net/e1000_osdep: replace schedule_timeout() with msleep

To: cramerj@xxxxxxxxx, john.ronciak@xxxxxxxxx, ganesh.venkatesan@xxxxxxxxx
Subject: Re: [PATCH 6/28] net/e1000_osdep: replace schedule_timeout() with msleep()
From: Nishanth Aravamudan <nacc@xxxxxxxxxx>
Date: Fri, 7 Jan 2005 13:37:44 -0800
Cc: kernel-janitors@xxxxxxxxxxxxxx, linux-net@xxxxxxxxxxxxxxx, netdev@xxxxxxxxxxx
In-reply-to: <20050107001931.GH3055@xxxxxxxxxx>
References: <20050107001931.GH3055@xxxxxxxxxx>
Sender: netdev-bounce@xxxxxxxxxxx
User-agent: Mutt/1.5.6+20040907i
On Thu, Jan 06, 2005 at 04:19:31PM -0800, Nishanth Aravamudan wrote:
> Hi,
> 
> Description: Use msleep() instead of schedule_timeout() to guarantee the task
> delays as expected.
> 
> Signed-off-by: Nishanth Aravamudan <nacc@xxxxxxxxxx>
> 
> 
> --- 2.6.10-v/drivers/net/e1000/e1000_osdep.h  2004-12-24 13:34:26.000000000 
> -0800
> +++ 2.6.10/drivers/net/e1000/e1000_osdep.h    2005-01-04 14:57:49.000000000 
> -0800
> @@ -46,8 +46,7 @@
>                               /* Don't mdelay in interrupt context! */ \
>                               BUG(); \
>                       } else { \
> -                             set_current_state(TASK_UNINTERRUPTIBLE); \
> -                             schedule_timeout((x * HZ)/1000 + 2); \
> +                             msleep(x);
>                       } } while(0)
>  /* Some workarounds require millisecond delays and are run during interrupt
>   * context.  Most notably, when establishing link, the phy may need tweaking

Please ignore this patch, as there is already a more correct one in the
2.6.10-kj patchset. Sorry for the noise.

Thanks,
Nish

<Prev in Thread] Current Thread [Next in Thread>