Re: [PATCH -V2 07/26] powerpc: Add size argument to pgtable_cache_add

["Aneesh Kumar K.V" <aneesh.kumar@linux.vnet.ibm.com>]

Paul Mackerras <paulus@samba.org> writes:

> On Wed, Mar 06, 2013 at 11:40:08AM +0530, Aneesh Kumar K.V wrote:
>> From: "Aneesh Kumar K.V" <aneesh.kumar@linux.vnet.ibm.com>
>> 
>> We will use this later with THP changes to request for pmd table of double the size.
>> THP code does PTE page allocation along with large page request and deposit them
>> for later use. This is to ensure that we won't have any failures when we split
>> huge pages to regular pages.
>> 
>> On powerpc we want to use the deposited PTE page for storing hash pte slot and
>> secondary bit information for the HPTEs. Hence we save them in the second half
>> of the pmd table.
>
> That much is fine, but...
>
>> -void pgtable_cache_add(unsigned shift, void (*ctor)(void *));
>> +extern void __pgtable_cache_add(unsigned shift, unsigned long table_size,
>> +				void (*ctor)(void *));
>>  void pgtable_cache_init(void);
>> +static inline void pgtable_cache_add(unsigned shift, void (*ctor)(void *))
>> +{
>> +	return __pgtable_cache_add(shift, sizeof(void *) << shift, ctor);
>> +}
>
> NAK to this (and in fact the rest of this patch).
>
> If you need twice the size, use the next slot along in pgtable_cache,
> which will give you the size you need, rather than breaking the
> relationship between slot index and allocation size.

That would mean we will have to do 

 static inline pmd_t *pmd_alloc_one(struct mm_struct *mm, unsigned long addr)
 {
-	return kmem_cache_alloc(PGT_CACHE(PMD_INDEX_SIZE),
+	return kmem_cache_alloc(PGT_CACHE(PMD_INDEX_SIZE + 1),
 				GFP_KERNEL|__GFP_REPEAT);
 }

is that ok ?

-aneesh