Re: vmalloc performance

[Minchan Kim]

Cced Nick.
He's Mr. Vmalloc.

On Wed, Apr 14, 2010 at 9:49 PM, Steven Whitehouse <swhiteho@xxxxxxxxxx> wrote:
>
> Since this didn't attract much interest the first time around, and at
> the risk of appearing to be talking to myself, here is the patch from
> the bugzilla to better illustrate the issue:
>
>
> diff --git a/mm/vmalloc.c b/mm/vmalloc.c
> index ae00746..63c8178 100644
> --- a/mm/vmalloc.c
> +++ b/mm/vmalloc.c
> @@ -605,8 +605,7 @@ static void free_unmap_vmap_area_noflush(struct
> vmap_area *va)
> Â{
> Â Â Â Âva->flags |= VM_LAZY_FREE;
> Â Â Â Âatomic_add((va->va_end - va->va_start) >> PAGE_SHIFT, &vmap_lazy_nr);
> - Â Â Â if (unlikely(atomic_read(&vmap_lazy_nr) > lazy_max_pages()))
> - Â Â Â Â Â Â Â try_purge_vmap_area_lazy();
> + Â Â Â try_purge_vmap_area_lazy();
> Â}
>
> Â/*
>
>
> Steve.
>
> On Mon, 2010-04-12 at 17:27 +0100, Steven Whitehouse wrote:
>> Hi,
>>
>> I've noticed that vmalloc seems to be rather slow. I wrote a test kernel
>> module to track down what was going wrong. The kernel module does one
>> million vmalloc/touch mem/vfree in a loop and prints out how long it
>> takes.
>>
>> The source of the test kernel module can be found as an attachment to
>> this bz: https://bugzilla.redhat.com/show_bug.cgi?id=581459
>>
>> When this module is run on my x86_64, 8 core, 12 Gb machine, then on an
>> otherwise idle system I get the following results:
>>
>> vmalloc took 148798983 us
>> vmalloc took 151664529 us
>> vmalloc took 152416398 us
>> vmalloc took 151837733 us
>>
>> After applying the two line patch (see the same bz) which disabled the
>> delayed removal of the structures, which appears to be intended to
>> improve performance in the smp case by reducing TLB flushes across cpus,
>> I get the following results:
>>
>> vmalloc took 15363634 us
>> vmalloc took 15358026 us
>> vmalloc took 15240955 us
>> vmalloc took 15402302 us
>>
>> So thats a speed up of around 10x, which isn't too bad. The question is
>> whether it is possible to come to a compromise where it is possible to
>> retain the benefits of the delayed TLB flushing code, but reduce the
>> overhead for other users. My two line patch basically disables the delay
>> by forcing a removal on each and every vfree.
>>
>> What is the correct way to fix this I wonder?
>>
>> Steve.
>>
>
> --
> To unsubscribe, send a message with 'unsubscribe linux-mm' in
> the body to majordomo@xxxxxxxxxx ÂFor more info on Linux MM,
> see: http://www.linux-mm.org/ .
> Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>
>



-- 
Kind regards,
Minchan Kim
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/