[RFC 00/10] implement alternative and much simpler id allocator

[Tejun Heo]

Hello, Matthew.

On Mon, Dec 12, 2016 at 05:35:17PM +0000, Matthew Wilcox wrote:
> I know the preload followed by preload_end looks wrong.  I don't
> think it's broken though.  If we get preempted, then the worst
> situation is that we'll end up with the memory we preallocated being
> allocated to somebody else.  Then the attempt to allocate memory can
> fail, and we'll return -EAGAIN, at which point all callers are
> supposed to return to the pre_get() stage.  Certainly that's what
> ida_simple_get() does.

Ah, right, ida_pre_get() doesn't have any protection against other
task allocating inbetween pre_get and the actual allocation, so it
should retry on failure.  Yeah, then the proposed preloading wouldn't
be wrong.  It'd be nice to explain what's going on tho.

> I'd definitely be open to changing the IDA API.  I know Kent had
> some thoughts on that including splitting the simple lock into a
> per-IDA lock.  His last work on it was here, I believe:
> 
> https://evilpiepirate.org/git/linux-bcache.git/log/?h=idr

Yeah, that was a big re-write, but for now I think it'd be nice to
replace ida's pre_get mechanism with something similar to idr's
preload so that they're more consistent.  There aren't that many
direct users of ida_pre_get(), so it shouldn't be too difficult to
change.

Thanks.

-- 
tejun