Re: [PATCH v2 1/2] vt: vt_ioctl: fix VT_DISALLOCATE freeing in-use virtual console

[Greg Kroah-Hartman <gregkh@linuxfoundation.org>]

On Thu, Mar 19, 2020 at 10:10:49PM -0700, Eric Biggers wrote:
> On Thu, Mar 19, 2020 at 08:36:28AM +0100, Jiri Slaby wrote:
> > On 18. 03. 20, 23:38, Eric Biggers wrote:
> > > --- a/drivers/tty/vt/vt.c
> > > +++ b/drivers/tty/vt/vt.c
> > > @@ -1102,6 +1102,9 @@ int vc_allocate(unsigned int currcons)	/* return 0 on success */
> > >  	tty_port_init(&vc->port);
> > >  	INIT_WORK(&vc_cons[currcons].SAK_work, vc_SAK);
> > >  
> > > +	/* if this wasn't the case, we'd have to implement port->ops.destruct */
> > > +	BUILD_BUG_ON(offsetof(struct vc_data, port) != 0);
> > > +
> > 
> > This is 3 lines, implementing destruct would be like 4-5 :)? Please
> > implement destruct instead.
> > 
> > Otherwise looks good.
> > 
> 
> Actually implementing destruct would be 12 lines, see below.  Remember there is
> no tty_port_operations defined yet so we'd have to define it just for destruct.
> 
> Do you still prefer it?
> 
> diff --git a/drivers/tty/vt/vt.c b/drivers/tty/vt/vt.c
> index ec34f1f5f3bb5..309a39197be0a 100644
> --- a/drivers/tty/vt/vt.c
> +++ b/drivers/tty/vt/vt.c
> @@ -1075,6 +1075,17 @@ static void visual_deinit(struct vc_data *vc)
>  	module_put(vc->vc_sw->owner);
>  }
>  
> +static void vc_port_destruct(struct tty_port *port)
> +{
> +	struct vc_data *vc = container_of(port, struct vc_data, port);
> +
> +	kfree(vc);
> +}
> +
> +static const struct tty_port_operations vc_port_ops = {
> +	.destruct = vc_port_destruct,
> +};
> +
>  int vc_allocate(unsigned int currcons)	/* return 0 on success */
>  {
>  	struct vt_notifier_param param;
> @@ -1100,11 +1111,9 @@ int vc_allocate(unsigned int currcons)	/* return 0 on success */
>  
>  	vc_cons[currcons].d = vc;
>  	tty_port_init(&vc->port);
> +	vc->port.ops = &vc_port_ops;
>  	INIT_WORK(&vc_cons[currcons].SAK_work, vc_SAK);
>  
> -	/* if this wasn't the case, we'd have to implement port->ops.destruct */
> -	BUILD_BUG_ON(offsetof(struct vc_data, port) != 0);
> -
>  	visual_init(vc, currcons, 1);
>  
>  	if (!*vc->vc_uni_pagedir_loc)


Yes, this is good to have, thanks for doing this.

greg k-h