On 01/12/2016 10:09 PM, Baoquan He wrote:
> Seems 0f:00.0 is not attached to a group so that it get a domain. Can
> you also paste your "lspci -t" and "lspci -vvv" output? Here 0f:00.0
> should be a pci bridge which need set the same domain with its
> peripheral 0000:0f:04.0.
>
> Please apply this patch and try again.
>
>
>>From 6740bc427ddfeecc4b557fcb3ceb0832f73ab33a Mon Sep 17 00:00:00 2001
> From: Baoquan He <bhe-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
> Date: Wed, 13 Jan 2016 11:02:28 +0800
> Subject: [PATCH] iommu/amd: Correct the wrong setting of alias DTE in
>   do_attach
>
> In below commit Joerg tried to set alias DTE when its peripheral
> is setting DTE. But there's a code bug here to wrongly set the
> alias DTE, correct it in this patch.
>
> commit e25bfb56ea7f046b71414e02f80f620deb5c6362
> Author: Joerg Roedel <jroedel-l3A5Bk7waGM@public.gmane.org>
> Date:   Tue Oct 20 17:33:38 2015 +0200
>
>      iommu/amd: Set alias DTE in do_attach/do_detach
>
> Signed-off-by: Baoquan He <bhe-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
> ---
>   drivers/iommu/amd_iommu.c | 2 +-
>   1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/iommu/amd_iommu.c b/drivers/iommu/amd_iommu.c
> index 8b2be1e..fc836f5 100644
> --- a/drivers/iommu/amd_iommu.c
> +++ b/drivers/iommu/amd_iommu.c
> @@ -1905,7 +1905,7 @@ static void do_attach(struct iommu_dev_data *dev_data,
>   	/* Update device table */
>   	set_dte_entry(dev_data->devid, domain, ats);
>   	if (alias != dev_data->devid)
> -		set_dte_entry(dev_data->devid, domain, ats);
> +		set_dte_entry(alias, domain, ats);
>
>   	device_flush_dte(dev_data);
>   }
>

First, attached are the 2 lspci outputs you asked for.

After the patch I was still unable to bring the system up but it was 
much better. 2 of the 8 Sata disks do not come up. If I remove those 
from my fstab the box comes up with the patch.

This MB has 8 internal Sata ports. Six hanging off the South-Bridge 
(SB950) and two more hanging off  Marvell 88SE9172 chips connected to 
the 990FX chipset.  I'm pretty sure the 2 disks that don't come up are 
the 2 hanging off the Marvell chips. I will have to take the box apart 
to verify and will if you need me to. And again, all works fine with the 
IOMMU disabled in the BIOS.

Regards
Mark