Re: [PATCH] Make GFP_DMA allocations w/o ZONE_DMA emit a warninginstead of failing

[Dmitry Eremin-Solenikov]

On 6/1/11, David Rientjes <rientjes@xxxxxxxxxx> wrote:
> On Wed, 1 Jun 2011, Dmitry Eremin-Solenikov wrote:
>
>> I've hit this with IrDA driver on PXA. Also I've seen the report regarding
>> other ARM platform (ep-something). Thus I've included Russell in the cc.
>>
>
> So you want to continue to allow the page allocator to return pages from
> anywhere, even when GFP_DMA is specified, just as though it was lowmem?

Yes and no. I'm asking for the grace period for the drivers authors to be able
to fix their code. After a grace period of one or two majors this permission
should be removed and your original patch should be effective.

> Why don't you actually address the problem with the driver you're
> complaining about with the patch below, which I already posted to you a
> few days ago?
>
> If this arm driver is going to be using GFP_DMA unconditionally, it better
> require CONFIG_ZONE_DMA for it to actually be meaningful until such time
> as it can be removed if it's truly not needed or generalized to only
> specific pieces of hardware.

No. This only workarounds the bug. And also a possible hundred of other bugs
in the PXA/etc. ARM drivers. Instead I'm asking for the way to
visualize all such
bugs.

Do you want to also add such workarounds to some PATA CF driver used on PXA?
To _any_ of the drivers allocating the GFP_DMA memory? Then CONFIG_ZONE_DMA
would serve no purpose. We can as well to drop that symbol. Believe me.

> ---
>  drivers/net/irda/Kconfig |    1 +
>  1 files changed, 1 insertions(+), 0 deletions(-)
>
> diff --git a/drivers/net/irda/Kconfig b/drivers/net/irda/Kconfig
> --- a/drivers/net/irda/Kconfig
> +++ b/drivers/net/irda/Kconfig
> @@ -374,6 +374,7 @@ config VIA_FIR
>  config PXA_FICP
>  	tristate "Intel PXA2xx Internal FICP"
>  	depends on ARCH_PXA && IRDA
> +	select ZONE_DMA
>  	help
>  	  Say Y or M here if you want to build support for the PXA2xx
>  	  built-in IRDA interface which can support both SIR and FIR.
>


-- 
With best wishes
Dmitry
--
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/