Re: [PATCH -mm 3/5] Hibernation: Correct definitions of some ioctls

["Rafael J. Wysocki" <rjw@sisk.pl>]

On Monday, 3 September 2007 05:25, Pavel Machek wrote:
> Hi!
> 
> > From: Rafael J. Wysocki <rjw@sisk.pl>
> > 
> > Three ioctl numbers belonging to the hibernation userland interface,
> > SNAPSHOT_ATOMIC_SNAPSHOT, SNAPSHOT_AVAIL_SWAP, SNAPSHOT_GET_SWAP_PAGE,
> > are defined in a wrong way (eg. not portable).  Provide new ioctl numbers for
> > these ioctls and mark the existing ones as deprecated.
> 
> It looks to me like
> 
> > @@ -149,12 +149,9 @@ struct resume_swap_area {
> >  #define SNAPSHOT_IOC_MAGIC	'3'
> >  #define SNAPSHOT_FREEZE			_IO(SNAPSHOT_IOC_MAGIC, 1)
> >  #define SNAPSHOT_UNFREEZE		_IO(SNAPSHOT_IOC_MAGIC, 2)
> > -#define SNAPSHOT_ATOMIC_SNAPSHOT	_IOW(SNAPSHOT_IOC_MAGIC, 3, void *)
> >  #define SNAPSHOT_ATOMIC_RESTORE		_IO(SNAPSHOT_IOC_MAGIC, 4)
> >  #define SNAPSHOT_FREE			_IO(SNAPSHOT_IOC_MAGIC, 5)
> >  #define SNAPSHOT_SET_IMAGE_SIZE  _IOW(SNAPSHOT_IOC_MAGIC, 6, unsigned long)
> 
> at least set_image_size is wrong - too: we do not use argument as
> unsigned long *, and we certainly don't write to it.

Yes, it is.  I think it should be _IO(SNAPSHOT_IOC_MAGIC, a_number)

I'll update the patch to cover this.

> > -#define SNAPSHOT_AVAIL_SWAP		_IOR(SNAPSHOT_IOC_MAGIC, 7, void *)
> > -#define SNAPSHOT_GET_SWAP_PAGE		_IOR(SNAPSHOT_IOC_MAGIC, 8, void *)
> >  #define SNAPSHOT_FREE_SWAP_PAGES	_IO(SNAPSHOT_IOC_MAGIC, 9)
> >  #define SNAPSHOT_SET_SWAP_FILE		_IOW(SNAPSHOT_IOC_MAGIC, 10, unsigned int)
> >  #define SNAPSHOT_S2RAM			_IO(SNAPSHOT_IOC_MAGIC, 11)
> > @@ -163,7 +160,10 @@ struct resume_swap_area {
> >  #define SNAPSHOT_GET_IMAGE_SIZE		_IOR(SNAPSHOT_IOC_MAGIC, 14, loff_t)
> >  #define SNAPSHOT_PLATFORM_SUPPORT	_IOR(SNAPSHOT_IOC_MAGIC, 15, int)

This one also should be _IO(...), BTW.

> >  #define SNAPSHOT_POWER_OFF		_IO(SNAPSHOT_IOC_MAGIC, 16)
> > -#define SNAPSHOT_IOC_MAXNR	16
> > +#define SNAPSHOT_CREATE_IMAGE		_IOW(SNAPSHOT_IOC_MAGIC, 17, int)
> > +#define SNAPSHOT_AVAIL_SWAP_SIZE	_IOR(SNAPSHOT_IOC_MAGIC, 18, -#loff_t)
> 
> Should this be _IOW, because we return that value to userspace?

No.  The direction is from the userland POV, AFAICS.

> > +#define SNAPSHOT_ALLOC_SWAP_PAGE	_IOR(SNAPSHOT_IOC_MAGIC, 19, loff_t)
> 
> Hmm, this looks like _IOW to me, too, plus it does put_user with u64
> (at least in my copy, may have been hacked too much).

In -rc5 it uses loff_t, and I believe that the direction is correct.

Greetings,
Rafael