[Jaunty SRU] LP#345710 [v2: description modified] iwl3945: cancel rfkill_poll with sync when?the module is exiting

[Huaxu Wan]

Hi Stefan,

The patch is fine for me. It works for 2.6.28 kernel. But, this bug is
not found with 2.6.30 or above kernel. This patch is not valuable to
2.6.30+, but 2.6.28 stable. I have no idea about whether it should be
send to upstream. Please do it if you want.

Thanks
Huaxu


On 11:49 Tue 30 Jun, Stefan Bader wrote:

> Hi Huaxu,
>
> we got now confirmed in the bug, that the upstream part alone does not 
> fix the problem.  prepared the following update to it. I added your 
> signed-off for now, is that ok? Should I send it upstream or do you want 
> to do that?
>
> Thanks,
> Stefan
>
>
> -- 
>
> When all other means of communication fail, try words!
>
>

> >From c64c59ea22b61a2c3f64dd269812b7ae5d3331ec Mon Sep 17 00:00:00 2001
> From: Stefan Bader <stefan.bader at canonical.com>
> Date: Fri, 26 Jun 2009 09:01:33 +0200
> Subject: [PATCH] iwl3945: Move cancel_delayed_work_sync to avoid lockup
> 
> BugLink: http://bugs.launchpad.net/ubuntu/+bug/345710
> 
> The issue of calling the rfkill workqueue on module remove had been tried to
> solve by commit 71d449b55abf5018d7c711b2b62abc0c083723c4.
> However the change from cancel_delayed_work to the sync version does not seem
> to be enough to prevent the lockup. Moving the cancelation down to a later
> time according to Huaxu Wan's version 2 patch.
> 
> Signed-off-by: Huaxu Wan <huaxu.wan at linux.intel.com>
> Signed-off-by: Stefan Bader <stefan.bader at canonical.com>
> Tested-by: Mcarni <mcarni at yahoo.it>
> ---
>  drivers/net/wireless/iwlwifi/iwl3945-base.c |    3 +--
>  1 files changed, 1 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/net/wireless/iwlwifi/iwl3945-base.c b/drivers/net/wireless/iwlwifi/iwl3945-base.c
> index cb9bd4c..e977af4 100644
> --- a/drivers/net/wireless/iwlwifi/iwl3945-base.c
> +++ b/drivers/net/wireless/iwlwifi/iwl3945-base.c
> @@ -4245,8 +4245,6 @@ static void __devexit iwl3945_pci_remove(struct pci_dev *pdev)
>  
>  	sysfs_remove_group(&pdev->dev.kobj, &iwl3945_attribute_group);
>  
> -	cancel_delayed_work_sync(&priv->rfkill_poll);
> -
>  	iwl3945_dealloc_ucode_pci(priv);
>  
>  	if (priv->rxq.bd)
> @@ -4258,6 +4256,7 @@ static void __devexit iwl3945_pci_remove(struct pci_dev *pdev)
>  
>  	/*netif_stop_queue(dev); */
>  	flush_workqueue(priv->workqueue);
> +	cancel_delayed_work_sync(&priv->rfkill_poll);
>  
>  	/* ieee80211_unregister_hw calls iwl3945_mac_stop, which flushes
>  	 * priv->workqueue... so we can't take down the workqueue
> -- 
> 1.5.4.3
> 


-- 
Thanks
Huaxu