Messages in this thread | | | Subject | Re: unregister ioctl32 conversion and modules. ioctl32 revisited. | From | Arnd Bergmann <> | Date | Wed, 15 Dec 2004 23:10:02 +0100 |
| |
"Michael S. Tsirkin" <mst@mellanox.co.il> schrieb am 15.12.2004, 19:21:09:
> But what if you really wanted to return -ENOIOCTLCMD?
It's not possible to return -ENOIOCTLCMD to user space, because it is a kernel internal value that was defined specifically the purpose of nesting ioctl handlers.
> And, the idea was to get rid of the hash eventually?
Oh, was it? I thought the idea was to get rid of register_ioctl32_conversion(). The use of the hash table may be not as efficient as possible, but it is not actually broken. The advantage of a global lookup mechanism is that many drivers don't even need to know about it.
Even if we decide to remove the hash at some later point (after the dynamic registration is gone), we definitely need a way to get back to the hashed functions from drivers in the meantime.
Arnd <>< - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |