Return-Path: <linux-kernel-owner+willy=40w.ods.org-S932213AbWEPWXf@vger.kernel.org>
Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand
	id S932213AbWEPWXf (ORCPT <rfc822;willy@w.ods.org>);
	Tue, 16 May 2006 18:23:35 -0400
Received: (majordomo@vger.kernel.org) by vger.kernel.org id S932218AbWEPWXf
	(ORCPT <rfc822;linux-kernel-outgoing>);
	Tue, 16 May 2006 18:23:35 -0400
Received: from hera.kernel.org ([140.211.167.34]:64130 "EHLO hera.kernel.org")
	by vger.kernel.org with ESMTP id S932214AbWEPWXd (ORCPT
	<rfc822;linux-kernel@vger.kernel.org>);
	Tue, 16 May 2006 18:23:33 -0400
To: linux-kernel@vger.kernel.org
From: Stephen Hemminger <shemminger@osdl.org>
Subject: Re: events/0 eats 100% cpu on core duo laptop
Date: Tue, 16 May 2006 15:22:56 -0700
Organization: OSDL
Message-ID: <20060516152256.34c2feea@localhost.localdomain>
References: <Pine.LNX.4.64.0605162002150.9606@tassadar.physics.auth.gr>
	<20060516122303.48b14dc1.akpm@osdl.org>
	<Pine.LNX.4.64.0605170041570.16352@tassadar.physics.auth.gr>
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
X-Trace: build.pdx.osdl.net 1147818176 29304 10.8.0.54 (16 May 2006 22:22:56 GMT)
X-Complaints-To: abuse@osdl.org
NNTP-Posting-Date: Tue, 16 May 2006 22:22:56 +0000 (UTC)
X-Newsreader: Sylpheed-Claws 2.1.0 (GTK+ 2.8.6; i486-pc-linux-gnu)
Sender: linux-kernel-owner@vger.kernel.org
X-Mailing-List: linux-kernel@vger.kernel.org
Content-Length: 2423
Lines: 59

On Wed, 17 May 2006 00:49:09 +0300 (EEST)
Dimitris Zilaskos <dzila@tassadar.physics.auth.gr> wrote:

> 
> 
>  	It seems that the issue occurs as soon as the nic is brought up:
> 
> Pid: 6, comm:             events/0
> EIP: 0060:[<c02e9e6f>] CPU: 0
> EIP is at __gm_phy_read+0x42/0x7b
>   EFLAGS: 00000293    Not tainted  (2.6.16.16 #1)
> EAX: 000004e8 EBX: 00002884 ECX: 7a7cd340 EDX: e0030000
> ESI: 00000003 EDI: 00002880 EBP: c1617340 DS: 007b ES: 007b
> CR0: 8005003b CR2: b7fd6ccb CR3: 005b4000 CR4: 000006d0
>   [<c02e9ed8>] gm_phy_read+0x30/0x5b
>   [<c02ec704>] sky2_phy_task+0x31/0x129
>   [<c012dffc>] run_workqueue+0x76/0xea
>   [<c02ec6d3>] sky2_phy_task+0x0/0x129
>   [<c012e1b2>] worker_thread+0x142/0x164
>   [<c0118fe0>] default_wake_function+0x0/0x12
>   [<c0118fe0>] default_wake_function+0x0/0x12
>   [<c012e070>] worker_thread+0x0/0x164
>   [<c01317b9>] kthread+0xb7/0xbd
>   [<c0131702>] kthread+0x0/0xbd
>   [<c01011a1>] kernel_thread_helper+0x5/0xb
> 
>  	The nic was not working properly at bootup (it was unable to 
> detect link,gain ip from dhcp) but I have discovered that adding an ifconfig
> eth0 up prior to launching dhcp worked:
> 
> sky2 eth0: enabling interface
> sky2 eth0: Link is up at 100 Mbps, full duplex, flow control tx
> sky2 eth0: disabling interface
> sky2 eth0: phy read timeout
> sky2 eth0: enabling interface
> sky2 eth0: phy read timeout
> sky2 eth0: speed/duplex mismatch<6>sky2 eth0: Link is up at 100 Mbps, full 
> duplex, flow control tx
> 
>  	Also when I shutdown/reboot sky2 eth0: phy read timeout messages flood
>   the console for a few seconds. The nic is detected as :
> 
> sky2 v0.15 addr 0xf0000000 irq 19 Yukon-EC Ultra(0xb4) rev 2
> 
>  	Best regards,

That's my driver.  I put a lot of fixes in for 2.6.17, one of which changes
the setup of the chip type used in this machine. The driver is significantly
changed from 2.6.16 so best bet is to just take the 2.6.17 latest version
and compile it on 2.6.16 (or update to 2.6.17). I got reports from Fujitsu E610
user that it now works fine.

Unfortunately, I don't have the hardware (or full specs for the chip) used
in that machine.  Maybe I need to get a MacBook.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/