Re: [GIT PULL FOR v4.18] R-Car VSP1 TLB optimisation

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Mon, May 28, 2018 at 07:03:05AM -0300, Mauro Carvalho Chehab wrote:
> I can't see how brx can be NULL. At the sequence of ifs:
> 
> 	if (pipe->num_inputs > 2)
>                 brx = &vsp1->bru->entity;
>         else if (pipe->brx && !drm_pipe->force_brx_release)
>                 brx = pipe->brx;
>         else if (!vsp1->bru->entity.pipe)
>                 brx = &vsp1->bru->entity;
>         else
>                 brx = &vsp1->brs->entity;
> 
> 
> The usage of brx = &(something) will always return a non NULL
> value[1].

That's not right.  It can be NULL if it's &foo->first_struct_member
and ->entity is the first struct member.  If it weren't the first struct
member then Smatch would say that brx was non-NULL.

> [1] It might be doing a NULL deref - with seems to be your concern
>     when you're talking about the case where vsp1->brs is NULL - but 
>     that's not what Smatch is complaining here.

If vsp1->bru were NULL, it wouldn't be a NULL dereference because it's
not dereferencing it; it's just taking the address.  On the path where
we do:
	else if (!vsp1->bru->entity.pipe)
		brx = &vsp1->bru->entity;

Then Smatch sees that vsp1->bru is dereferenced and marks "brx" as
non-NULL.

regards,
dan carpenter




[Index of Archives]     [Linux Input]     [Video for Linux]     [Gstreamer Embedded]     [Mplayer Users]     [Linux USB Devel]     [Linux Audio Users]     [Linux Kernel]     [Linux SCSI]     [Yosemite Backpacking]

  Powered by Linux