[PATCH] [xkb] Fix possible NULL dereference in XkbFlushLedEvents()

[Ian Romanick]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Tomas Carnecky wrote:
> Through some code paths it is possible that NULL is being passed in the
> 'ed' parameter to XkbFlushLedEvents(). Make sure we don't pass it along
> to bzero().
> 
> Signed-off-by: Tomas Carnecky <tom at dbservice.com>
> ---
>  xkb/xkbLEDs.c |    3 ++-
>  1 files changed, 2 insertions(+), 1 deletions(-)
> 
> diff --git a/xkb/xkbLEDs.c b/xkb/xkbLEDs.c
> index 59cdba4..dfdd5a2 100644
> --- a/xkb/xkbLEDs.c
> +++ b/xkb/xkbLEDs.c
> @@ -750,7 +750,8 @@ XkbFlushLedEvents(	DeviceIntPtr			dev,
>  	    XkbDDXUpdateDeviceIndicators(dev,sli,sli->effectiveState);
>  	XkbSendExtensionDeviceNotify(dev,cause->client,ed);
>      }
> -    bzero((char *)ed,sizeof(XkbExtensionDeviceNotify));
> +    if (ed)
> +	bzero((char *)ed,sizeof(XkbExtensionDeviceNotify));

Since the preceding if-statement is "if (ed && (ed->reason))", I think
it would be more clear to restructure the code as:

    if (ed) {
        if (ed->reason)) {
            if ((dev!=kbd)&&(ed->reason&XkbXI_IndicatorStateMask))
                XkbDDXUpdateDeviceIndicators(dev,sli,sli->effectiveState);
            XkbSendExtensionDeviceNotify(dev,cause->client,ed);
        }

        bzero((char *)ed,sizeof(XkbExtensionDeviceNotify));
    }

Signed-off-by: Ian Romanick <ian.d.romanick at intel.com>

>      return;
>  }
>  
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iEYEARECAAYFAksqfssACgkQX1gOwKyEAw8WAACdEfWfUrVNnIKwCjDJzwrcyHVB
xA8AnR6eJ6nbHGEgeodaVdtLX+5AIa4d
=NlaA
-----END PGP SIGNATURE-----