2012/10/24 Christoph Lameter <cl@xxxxxxxxx>:
> On Sun, 21 Oct 2012, JoonSoo Kim wrote:
>
>> 2012/10/19 Christoph Lameter <cl@xxxxxxxxx>:
>>
>> > @@ -693,20 +657,19 @@ static inline struct array_cache *cpu_ca
>> > static inline struct kmem_cache *__find_general_cachep(size_t size,
>> > gfp_t gfpflags)
>> > {
>> > - struct cache_sizes *csizep = malloc_sizes;
>> > + int i;
>> >
>> > #if DEBUG
>> > /* This happens if someone tries to call
>> > * kmem_cache_create(), or __kmalloc(), before
>> > * the generic caches are initialized.
>> > */
>> > - BUG_ON(malloc_sizes[INDEX_AC].cs_cachep == NULL);
>> > + BUG_ON(kmalloc_caches[INDEX_AC] == NULL);
>> > #endif
>> > if (!size)
>> > return ZERO_SIZE_PTR;
>> >
>> > - while (size > csizep->cs_size)
>> > - csizep++;
>> > + i = kmalloc_index(size);
>>
>> Above kmalloc_index(size) is called with arbitrary size, therefore it
>> cannot be folded.
>
> The size is passed into an inline function that is folded and therefore
> the kmalloc_index function can also be folded if the size passed into
> __find_general_cachep was constant.
>
__find_general_cachep() is called by __do_kmalloc().
And __do_kmalloc() is called by __kmalloc().
__kmalloc() is called by kmalloc() when buildin_constant_p is failed.
Therefore __find_general_cachep() is called with arbitrary size value
and kmalloc_index() cannot be folded.
But, I think it doesn't matter, because CK2[15/15] also solve this problem.
Thanks.
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@xxxxxxxxx. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>