Re: [RFC PATCH v1 1/3] mm: teach mm by current context info to not doI/O during memory allocation

[Ming Lei]

On Wed, Oct 17, 2012 at 1:14 PM, Kamezawa Hiroyuki
<kamezawa.hiroyu@xxxxxxxxxxxxxx> wrote:
>
> I think the idea is reasonable. I have a request.

Thanks for your comment.

>
> In current implemententation of vmscan.c, it seems sc.may_writepage, sc.may_swap
> are handled independent from gfp_mask.
>
> So, could you drop changes from this patch and handle these flags in another patch
> if these flags should be unset if ~GFP_IOFS ?

OK, I agree. In theory,  mm should make sure no I/O is involved if
memory allocation
users passes ~GFP_IOFS.

>
> I think try_to_free_page() path's sc.may_xxxx should be handled in the same way.

Yes, alloc_page_buffers() and dma_alloc_from_contiguous may drop into
the path, so gfp flag should be changed in try_to_free_page() too.


Thanks,
--
Ming Lei
--
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/