Re: 2.6.2-rc1-mm1

[Greg KH]

On Thu, Jan 22, 2004 at 03:19:43PM +0000, viro@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx wrote:
> On Thu, Jan 22, 2004 at 11:03:42AM +0000, Christoph Hellwig wrote:
> > > sysfs-class-06-raw.patch
> > >   From: Greg KH <greg@xxxxxxxxx>
> > >   Subject: [PATCH] add sysfs class support for raw devices [06/10]
> > 
> > This one exports get_gendisk, which is a no-go.
> 
> Moreover, it obviously leaks references to struct gendisk _and_ changes
> semantics of RAW_SETBIND in incompatible way.
> 
> Consider that vetoed.  And yes, get_gendisk() issue alone would be enough.

Yes, I realize this patch isn't quite sane yet, that's why it's still in
the -mm tree :)

> Greg, please, RTFS to see at which point do we decide which driver will
> be used by raw device.  It's _not_ RAW_SETBIND, it's open().  So where
> your symlink should point is undecided until the same point.

I don't care about which driver is used by the raw device, I care about
which block device the raw device is "bound" to.  That happens at
RAW_SETBIND time, right?  We do this in the line:
	rawdev->binding = bdget(dev);

Hm, wait, are you saying that at open() time we actually bind the char
device to the block device, and that struct block_device can change
between RAW_SETBIND and open() time due to hotpluggable devices?

Or am I missing something else here?

thanks,

greg k-h
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/