linux-kernel - Re: [PATCH v3] firewire: Fix ohci free

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <20130217094100.7a5d17e3@stein>
Date:	Sun, 17 Feb 2013 09:41:00 +0100
From:	Stefan Richter <stefanr@...6.in-berlin.de>
To:	Mark Einon <mark.einon@...il.com>
Cc:	linux1394-devel@...ts.sourceforge.net,
	linux-kernel@...r.kernel.org, linux-pm@...r.kernel.org,
	stern@...land.harvard.edu
Subject: Re: [PATCH v3] firewire: Fix ohci free_irq() warning

On Feb 05 Mark Einon wrote:
> This patch fixes the kernel warning below, generated when putting an
> MSI MS-1727 GT740 laptop into suspend mode. The call sequence in this
> case calls free_irq() twice, once in pci_suspend() and once then in
> pci_remove().
> 
> The fix breaks up the ohci_enable() call into four separate calls -
> ohci_enable_regs(), ohci_create_irq(), ohci_enable_irq() and
> ohci_enable_run(). The original call sequence of ohci_enable() is replaced
> by the four calls, but the ohci_enable() called from pci_resume() is
> replaced by three of the calls, missing out ohci_create_irq().
> 
> Finally, a new call, ohci_disable_irq(), replaces the free_irq() in
> pci_suspend() which sets the irq mask to zero.
[...]

Hi,
I haven't forgotten about this patch; just didn't fully check it yet
because I still wanted to look whether there is a more minimalistic
way.  Perhaps I'll post something in the next days.

> @@ -3766,7 +3811,7 @@ static int pci_suspend(struct pci_dev *dev, pm_message_t state)
>  	int err;
>  
>  	software_reset(ohci);
> -	free_irq(dev->irq, ohci);
> +	ohci_disable_irq(ohci);
>  	pci_disable_msi(dev);
>  	err = pci_save_state(dev);
>  	if (err) {
> @@ -3802,10 +3847,13 @@ static int pci_resume(struct pci_dev *dev)
>  		reg_write(ohci, OHCI1394_GUIDHi, (u32)(ohci->card.guid >> 32));
>  	}
>  
> -	err = ohci_enable(&ohci->card, NULL, 0);
> +	err = ohci_enable_regs(&ohci->card, NULL, 0);
>  	if (err)
>  		return err;
>  
> +	ohci_enable_irq(ohci);
> +	ohci_enable_run(ohci);
> +
>  	ohci_resume_iso_dma(ohci);
>  
>  	return 0;

Looks like pci_{en,dis}able_msi are now unbalanced.  I suppose we better
place them immediately around {request,free}_irq.
-- 
Stefan Richter
-=====-===-= --=- =---=
http://arcgraph.de/sr/
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/