Re: [PATCH] sound: depend on ZONE_DMA

[Takashi Iwai <tiwai@suse.de>]

On Mon, 16 Nov 2015 09:58:28 +0100,
Boaz Harrosh wrote:
> 
> On 11/16/2015 09:40 AM, Takashi Iwai wrote:
> > On Sun, 15 Nov 2015 11:53:11 +0100,
> > Boaz Harrosh wrote:
> >>
> >> On 11/12/2015 10:38 PM, Takashi Iwai wrote:
> >>> On Thu, 12 Nov 2015 21:13:57 +0100,
> >>> Dan Williams wrote:
> >>>>
> >>>> There are several sound drivers that 'select ZONE_DMA'.  This is
> >>>> backwards as ZONE_DMA is an architecture capability exported to drivers.
> >>>> Switch the polarity of the dependency to disable these drivers when the
> >>>> architecture does not support ZONE_DMA.  This was discovered in the
> >>>> context of testing/enabling devm_memremap_pages() which depends on
> >>>> ZONE_DEVICE.  ZONE_DEVICE in turn depends on !ZONE_DMA.
> >>>
> >>> Makes sense.  I applied it now, thanks.
> >>>
> >>
> >> Please add:
> >> CC: Stable Tree <stable@vger.kernel.org>
> > 
> > Sorry, too late, already merged.
> > 
> > 
> > Takashi
> > 
> 
> Hi Greg
> 
> Please include the mainline patch:
> 	[2db1a57] ALSA: pci: depend on ZONE_DMA (by Dan Williams)
> 
> To the stable tree for v4.3.X Kernel.
> 
> This patch is needed for proper operation of the 4.3 pmem.ko driver. Long
> story, but without this patch the persistent-memory will not be able to
> work with the new 4.3 support of page-struct which is needed if we want to
> RDMA and/or IO directly to persistent memory.
> [Is to do with the new ZONE_DEVICE and too many ZONE(s) if ZONE_DMA is enabled]

Well, it's not exactly true: you just need to deselect some drivers to
allow CONFIG_ZONE_DMA to be disabled.  It's merely another side of
coin.

I don't mind including this to 4.3, though, but just want to correct
the statement.


Takashi