Re: drivers/parport/procfs.c:267:18-28: WARNING use flexible-array member instead (https://www.kernel.org/doc/html/latest/process/deprecated.html#zero-length-and-one-element-arrays) (fwd)

[Joel Granados]

On Sat, Nov 02, 2024 at 10:05:24PM +0100, Julia Lawall wrote:
> If device_dir is meant to be a flexible array, use [] instead or [1].
I don't think it is meant as a flexible array. I'll try to contextualize what is
happening and suggest a way forward.

> 
> julia
> 
> ---------- Forwarded message ----------
> Date: Sun, 3 Nov 2024 04:56:55 +0800
> From: kernel test robot <lkp@xxxxxxxxx>
> To: oe-kbuild@xxxxxxxxxxxxxxx
> Cc: lkp@xxxxxxxxx, Julia Lawall <julia.lawall@xxxxxxxx>

...

> vim +267 drivers/parport/procfs.c
> 
> ^1da177e4c3f41 Linus Torvalds 2005-04-16  257
> ^1da177e4c3f41 Linus Torvalds 2005-04-16  258
> ^1da177e4c3f41 Linus Torvalds 2005-04-16  259  struct parport_sysctl_table {
> 37e9981e33e4d3 Joel Granados  2023-06-16  260  	struct ctl_table_header *port_header;
> 37e9981e33e4d3 Joel Granados  2023-06-16  261  	struct ctl_table_header *devices_header;
> 0829381e481ab2 Joel Granados  2023-10-02  262  #ifdef CONFIG_PARPORT_1284
> 0829381e481ab2 Joel Granados  2023-10-02  263  	struct ctl_table vars[10];
> 0829381e481ab2 Joel Granados  2023-10-02  264  #else
> 0829381e481ab2 Joel Granados  2023-10-02  265  	struct ctl_table vars[5];
> 0829381e481ab2 Joel Granados  2023-10-02  266  #endif /* IEEE 1284 support */
> 0829381e481ab2 Joel Granados  2023-10-02 @267  	struct ctl_table device_dir[1];
> ^1da177e4c3f41 Linus Torvalds 2005-04-16  268  };
> ^1da177e4c3f41 Linus Torvalds 2005-04-16  269

What we were doing was just to remove the sentinel from that device_dir array.
As you can it was originally sized at 2 (with the second element left empty to
signify a sentinel). This is the diff:
``` diff
  diff --git a/drivers/parport/procfs.c b/drivers/parport/procfs.c
   index 4e5b972c3e26..532d5cbbd344 100644
   --- a/drivers/parport/procfs.c
   +++ b/drivers/parport/procfs.c
   @@ -259,8 +259,12 @@ PARPORT_MAX_SPINTIME_VALUE;
    struct parport_sysctl_table {
          struct ctl_table_header *port_header;
          struct ctl_table_header *devices_header;
   -      struct ctl_table vars[12];
   -      struct ctl_table device_dir[2];
   +#ifdef CONFIG_PARPORT_1284
   +      struct ctl_table vars[10];
   +#else
   +      struct ctl_table vars[5];
   +#endif /* IEEE 1284 support */
   +      struct ctl_table device_dir[1];
    };
```

So the array was at the end with a size to begin with. But I don't think that
its original intention was to be a flexible array. I'm not sure though, I'm
ccing the parport maintainers in hope that they can shed more light on this.

Notice that I prefer device_dir to be an array, so it can be passed to the
sysctl register call as is

Best

> 
> -- 
> 0-DAY CI Kernel Test Service
> https://github.com/intel/lkp-tests/wiki

-- 

Joel Granados