On Wed, Jul 11, 2012 at 12:24:41AM +0900, JoonSoo Kim wrote:
> > That would be functionally similar to your patch but it will preserve git
> > blame, churn less code and be harder to make mistakes with in the unlikely
> > event a third call to alloc_pages_direct_compact is ever added.
>
> Your sug
2012/7/10 Mel Gorman :
> You say that invoking the function is very costly. I agree that a function
> call with that many parameters is hefty but it is also in the slow path of
> the allocator. For order-0 allocations we are about to enter direct reclaim
> where I would expect the cost far exceeds
On Sun, Jul 08, 2012 at 11:33:14AM +0900, JoonSoo Kim wrote:
> 2012/7/7 David Rientjes :
> > On Sat, 7 Jul 2012, Joonsoo Kim wrote:
> >
> >> __alloc_pages_direct_compact has many arguments so invoking it is very
> >> costly.
> >> And in almost invoking case, order is 0, so return immediately.
> >>
On Mon, 9 Jul 2012 23:13:50 +0900
JoonSoo Kim wrote:
> >> In my kernel image, __alloc_pages_direct_compact() is not inlined by gcc.
My gcc-4.4.4 doesn't inline it either.
> I think __alloc_pages_direct_compact() can't be inlined by gcc,
> because it is so big and is invoked two times in __alloc
On Mon, 9 Jul 2012, JoonSoo Kim wrote:
> I think __alloc_pages_direct_compact() can't be inlined by gcc,
> because it is so big and is invoked two times in __alloc_pages_nodemask().
>
We could fix that by doing
diff --git a/mm/page_alloc.c b/mm/page_alloc.c
--- a/mm/page_alloc.c
+++ b/mm/page_a
2012/7/9 David Rientjes :
> On Sun, 8 Jul 2012, JoonSoo Kim wrote:
>
>> >> __alloc_pages_direct_compact has many arguments so invoking it is very
>> >> costly.
>> >> And in almost invoking case, order is 0, so return immediately.
>> >>
>> >
>> > If "zero cost" is "very costly", then this might mak
On Sun, 8 Jul 2012, JoonSoo Kim wrote:
> >> __alloc_pages_direct_compact has many arguments so invoking it is very
> >> costly.
> >> And in almost invoking case, order is 0, so return immediately.
> >>
> >
> > If "zero cost" is "very costly", then this might make sense.
> >
> > __alloc_pages_dire
2012/7/7 David Rientjes :
> On Sat, 7 Jul 2012, Joonsoo Kim wrote:
>
>> __alloc_pages_direct_compact has many arguments so invoking it is very
>> costly.
>> And in almost invoking case, order is 0, so return immediately.
>>
>
> If "zero cost" is "very costly", then this might make sense.
>
> __all
>>
>> >> And in almost invoking case, order is 0, so return immediately.
>> >
>> > You can't make sure it.
>>
>> Okay.
>>
>> >>
>> >> Let's not invoke it when order 0
>> >
>> > Let's not ruin git blame.
>>
>> Hmm...
>> When I do git blame, I can't find anything related to this.
>
> I mean if we mer
On Sat, 7 Jul 2012, Joonsoo Kim wrote:
> __alloc_pages_direct_compact has many arguments so invoking it is very costly.
> And in almost invoking case, order is 0, so return immediately.
>
If "zero cost" is "very costly", then this might make sense.
__alloc_pages_direct_compact() is inlined by g
On Sat, Jul 07, 2012 at 01:58:24AM +0900, JoonSoo Kim wrote:
> 2012/7/7 Minchan Kim :
> > Hi Joonsoo,
> >
> > On Sat, Jul 07, 2012 at 12:28:41AM +0900, Joonsoo Kim wrote:
> >> __alloc_pages_direct_compact has many arguments so invoking it is very
> >> costly.
> >
> > It's already slow path so it's
2012/7/7 Minchan Kim :
> Hi Joonsoo,
>
> On Sat, Jul 07, 2012 at 12:28:41AM +0900, Joonsoo Kim wrote:
>> __alloc_pages_direct_compact has many arguments so invoking it is very
>> costly.
>
> It's already slow path so it's pointless for such optimization.
I know this is so minor optimization.
But
Hi Joonsoo,
On Sat, Jul 07, 2012 at 12:28:41AM +0900, Joonsoo Kim wrote:
> __alloc_pages_direct_compact has many arguments so invoking it is very costly.
It's already slow path so it's pointless for such optimization.
> And in almost invoking case, order is 0, so return immediately.
You can't m
__alloc_pages_direct_compact has many arguments so invoking it is very costly.
And in almost invoking case, order is 0, so return immediately.
Let's not invoke it when order 0
Signed-off-by: Joonsoo Kim
diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index 6092f33..f4039aa 100644
--- a/mm/page_a
14 matches
Mail list logo
