Re: Slab: Remove kmem_cache_t

[Nick Piggin <nickpiggin@yahoo.com.au>]

(sorry about the clock skew)

Nick Piggin wrote:
> Christoph Lameter wrote:
> 
>>
>> kmem_cache_t would require a declaration. struct kmem_cache * can be 
>> used without a prior declaration in include files. Please review the 
>> earlier discussion on linux-mm regarding the removal of the global 
>> slab caches from <linux/slab.h>.
>>
>> Frankly the maintenance of the opaque type here has caused us enough 
>> grief over the years. I would like to get rid of it in the future and 
>> declare the contents of struct kmem_cache in slab.h. That will allow 
>> us to simplify the slab bootstrap and make it easier to understand. 
>> One reason slab bootstrap is so complex because one cannot simple do a 
>> static declaration of a struct kmem_cache and start off with it. See 
>> the earlier discussion with Matt Mackall on the slabifier design.
> 
> 
> That's all fine, we're not talking about any of the mechanism to how any
> of that works, just what the type is called. So what exactly is wrong with
> a kmem_cache_t declaration in include files, then?


Also note that when I say opaque type, I don't mean that you should leave
the declaration of struct kmem_cache in mm/slab.c.

I just mean that this patch seems like pointless churn, and a typedef is
reasonable if not desirable in slab API users where kmem_cache_t *is* an
opaque type (ie. they have no business accessing any struct members).

-- 
SUSE Labs, Novell Inc.
Send instant messages to your online friends http://au.messenger.yahoo.com 

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>