Re: [PATCH] nbd driver for 2.5+: fix for module removal & new block device layer

[viro@parcelfarce.linux.theplanet.co.uk]

On Sat, Jun 21, 2003 at 03:48:56PM -0600, Lou Langholtz wrote:
> This patch prevents memory corruption from "rmmod nbd" with the existing 
> 2.5.72 (and earlier) nbd driver. It does this by updating the nbd driver 
> to the new block layer requirement that every disk has its own 
> request_queue structure. This is the first of a series of patchlets 
> designed to break down the essential changes I proposed in my last 
> "enormous" patch. Note that another patchlet will make the whole 
> allocation of per nbd_device memory be dynamic (rather than staticly 
> tied to MAX_NBD). Please try out this patch and let me know how nbd is 
> working for you before versus after. With any luck, some of these 
> smaller patch breakdowns can actually see there way into new kernel 
> releases. Thanks.

	a) you don't have to have queue per device.  It often does make
sense (for nbd it's almost certainly a win), but it's not required.

	b) you definitely don't have to use separate queue locks.  The
thing will work fine with spinlock being shared and I doubt that there
will be any noticable extra contention.

	c) please, make allocation of queue dynamic _and_ separate from
any other allocated objects.