Re: [PATCH] snd/hda: fix use-after-free after module unload

Re: [PATCH] snd/hda: fix use-after-free after module unload

[Takashi Iwai]

On Sat, 09 Jul 2016 16:38:57 +0200,
Peter Wu wrote:
> 
> register_vga_switcheroo() sets the PM ops from the hda structure which
> is freed later in azx_free. Make sure that these ops are cleared.
> 
> Caught by KASAN.
> 
> Fixes: 246efa4a072f ("snd/hda: add runtime suspend/resume on optimus support (v4)")
> Signed-off-by: Peter Wu <peter@lekensteyn.nl>
> ---
>  sound/pci/hda/hda_intel.c | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/sound/pci/hda/hda_intel.c b/sound/pci/hda/hda_intel.c
> index 94089fc..a339066 100644
> --- a/sound/pci/hda/hda_intel.c
> +++ b/sound/pci/hda/hda_intel.c
> @@ -1219,6 +1219,7 @@ static int azx_free(struct azx *chip)
>  			snd_hda_unlock_devices(&chip->bus);
>  		if (hda->vga_switcheroo_registered)
>  			vga_switcheroo_unregister_client(chip->pci);
> +        vga_switcheroo_fini_domain_pm_ops(chip->card->dev);

The domain pm ops is set only when hda->vga_switcheroo_registered flag
is set.  So the call should be in the previous if block.

Also, the indentation looks wrong.  Please use the correct
indentation.

Could you resubmit with these fixes?


thanks,

Takashi

Re: [PATCH] snd/hda: fix use-after-free after module unload

[Peter Wu]

On Mon, Jul 11, 2016 at 12:42:27PM +0200, Takashi Iwai wrote:
> On Sat, 09 Jul 2016 16:38:57 +0200,
> Peter Wu wrote:
> > 
> > register_vga_switcheroo() sets the PM ops from the hda structure which
> > is freed later in azx_free. Make sure that these ops are cleared.
> > 
> > Caught by KASAN.
> > 
> > Fixes: 246efa4a072f ("snd/hda: add runtime suspend/resume on optimus support (v4)")
> > Signed-off-by: Peter Wu <peter@lekensteyn.nl>
> > ---
> >  sound/pci/hda/hda_intel.c | 1 +
> >  1 file changed, 1 insertion(+)
> > 
> > diff --git a/sound/pci/hda/hda_intel.c b/sound/pci/hda/hda_intel.c
> > index 94089fc..a339066 100644
> > --- a/sound/pci/hda/hda_intel.c
> > +++ b/sound/pci/hda/hda_intel.c
> > @@ -1219,6 +1219,7 @@ static int azx_free(struct azx *chip)
> >  			snd_hda_unlock_devices(&chip->bus);
> >  		if (hda->vga_switcheroo_registered)
> >  			vga_switcheroo_unregister_client(chip->pci);
> > +        vga_switcheroo_fini_domain_pm_ops(chip->card->dev);
> 
> The domain pm ops is set only when hda->vga_switcheroo_registered flag
> is set.  So the call should be in the previous if block.

Yes that would be cleaner, will do that.

> Also, the indentation looks wrong.  Please use the correct
> indentation.

Noticed too late that the editor config was wrong on the testing
machine.

> Could you resubmit with these fixes?

I will, thanks for the feedback!
-- 
Kind regards,
Peter Wu
https://lekensteyn.nl

[PATCH v2] snd/hda: fix use-after-free after module unload

[Peter Wu]

register_vga_switcheroo() sets the PM ops from the hda structure which
is freed later in azx_free. Make sure that these ops are cleared.

Caught by KASAN, initially noticed due to a general protection fault.

Fixes: 246efa4a072f ("snd/hda: add runtime suspend/resume on optimus support (v4)")
Signed-off-by: Peter Wu <peter@lekensteyn.nl>
---
Maybe Cc stable?
---
 sound/pci/hda/hda_intel.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/sound/pci/hda/hda_intel.c b/sound/pci/hda/hda_intel.c
index 94089fc..4aeed98 100644
--- a/sound/pci/hda/hda_intel.c
+++ b/sound/pci/hda/hda_intel.c
@@ -1217,8 +1217,10 @@ static int azx_free(struct azx *chip)
 	if (use_vga_switcheroo(hda)) {
 		if (chip->disabled && hda->probe_continued)
 			snd_hda_unlock_devices(&chip->bus);
-		if (hda->vga_switcheroo_registered)
+		if (hda->vga_switcheroo_registered) {
 			vga_switcheroo_unregister_client(chip->pci);
+			vga_switcheroo_fini_domain_pm_ops(chip->card->dev);
+		}
 	}
 
 	if (bus->chip_init) {
-- 
2.9.0

Re: [PATCH v2] snd/hda: fix use-after-free after module unload

[Takashi Iwai]

On Mon, 11 Jul 2016 19:51:06 +0200,
Peter Wu wrote:
> 
> register_vga_switcheroo() sets the PM ops from the hda structure which
> is freed later in azx_free. Make sure that these ops are cleared.
> 
> Caught by KASAN, initially noticed due to a general protection fault.
> 
> Fixes: 246efa4a072f ("snd/hda: add runtime suspend/resume on optimus support (v4)")
> Signed-off-by: Peter Wu <peter@lekensteyn.nl>
> ---
> Maybe Cc stable?

Yes, I applied with Cc to stable now.


thanks,

Takashi