On Tue, Oct 19, 2021 at 08:32:57AM +0200, Sebastien Marie wrote:
> Hi,
>
> Simple online diff for properly nullify v_data (which is `void *`)
> with NULL instead of 0.
>
> Comments or OK ?
> --
> Sebastien Marie
>
There are many others along those lines in the kernel, for example
sparse complains about these in vfs_subr.c
/sys/kern/vfs_subr.c:274:64: warning: Using plain integer as NULL pointer
/sys/kern/vfs_subr.c:275:64: warning: Using plain integer as NULL pointer
/sys/kern/vfs_subr.c:430:32: warning: Using plain integer as NULL pointer
/sys/kern/vfs_subr.c:467:22: warning: Using plain integer as NULL pointer
/sys/kern/vfs_subr.c:533:50: warning: Using plain integer as NULL pointer
/sys/kern/vfs_subr.c:1150:77: warning: Using plain integer as NULL pointer
/sys/kern/vfs_subr.c:1430:42: warning: Using plain integer as NULL pointer
/sys/kern/vfs_subr.c:1483:19: warning: Using plain integer as NULL pointer
/sys/kern/vfs_subr.c:1752:25: warning: Using plain integer as NULL pointer
If this is something worth changing should it be a larger diff?
>
> blob - 1f7409235f4696765c6b8b22e65d953b1d6e5100
> blob + 8cb011e9c3e48fcc34b8ce62974cafd430bb4e6a
> --- sys/kern/vfs_subr.c
> +++ sys/kern/vfs_subr.c
> @@ -469,7 +469,7 @@ getnewvnode(enum vtagtype tag, struct mount *mp, const
> panic("%s: free vnode %p isn't lock free", __func__, vp);
> #endif
> vp->v_usecount = 1;
> - vp->v_data = 0;
> + vp->v_data = NULL;
> return (0);
> }
>
>
>
