Re: [PATCH v2 1/2] pci: Add support for multiple DMA aliases

[Bjorn Helgaas <helgaas@kernel.org>]

On Tue, Jan 26, 2016 at 04:42:23PM -0700, Alex Williamson wrote:
> On Tue, 2016-01-26 at 17:31 -0600, Bjorn Helgaas wrote:
> > Hi Jacek,
> > 
> > When you post updated versions, please post the whole series with a
> > new version number.  I think there are three postings of [v2 1/2], and
> > I think they're different.  In any case, it's hard for me to keep them
> > straight.
> > 
> > On Tue, Jan 26, 2016 at 06:12:15PM +0100, Jacek Lawrynowicz wrote:
> > > This patch solves IOMMU support issues with PCIe non-transparent bridges
> > > that use Requester ID look-up tables (LUT), e.g. PEX8733. Before exiting
> > > the bridge, packet's RID is rewritten according to LUT programmed by
> > > a driver. Modified packets are then passed to a destination bus and
> > > processed upstream. The problem is that such packets seem to come from
> > > non-existent nodes that are hidden behind NTB and are not discoverable
> > > by a destination node, so IOMMU discards them. Adding DMA alias for a
> > > given LUT entry allows IOMMU to create a proper mapping that enables
> > > inter-node communication.
> > > 
> > > The current DMA alias implementation supports only single alias, so it's
> > > not possible to connect more than two nodes when IOMMU is enabled. This
> > > implementation enables all possible aliases on a given bus (256) that
> > > are stored in a bitset. Alias devfn is directly translated to a bit
> > > number. The bitset is not allocated for devices that have no need for
> > > DMA aliases.
> > > 
> > > More details can be found in following article:
> > > http://www.plxtech.com/files/pdf/technical/expresslane/RTC_Enabling%20MulitHostSystemDesigns.pdf
> > > 
> > > Signed-off-by: Jacek Lawrynowicz <jacek.lawrynowicz@intel.com>
> > > Acked-By: David Woodhouse <David.Woodhouse@intel.com>
> > > ---
> > >  drivers/iommu/iommu.c | 11 ++++-------
> > >  drivers/pci/pci.c     | 27 +++++++++++++++++++++++++++
> > >  drivers/pci/pci.h     |  2 ++
> > >  drivers/pci/probe.c   |  1 +
> > >  drivers/pci/quirks.c  | 15 ++++++---------
> > >  drivers/pci/search.c  | 14 +++++++++-----
> > >  include/linux/pci.h   | 14 +++++++-------
> > >  7 files changed, 56 insertions(+), 28 deletions(-)
> > > 
> > > diff --git a/drivers/iommu/iommu.c b/drivers/iommu/iommu.c
> > > index 0e3b009..93a89ed 100644
> > > --- a/drivers/iommu/iommu.c
> > > +++ b/drivers/iommu/iommu.c
> > > @@ -660,8 +660,8 @@ static struct iommu_group *get_pci_function_alias_group(struct pci_dev
> > > *pdev,
> > >  }
> > >  
> > >  /*
> > > - * Look for aliases to or from the given device for exisiting groups.  The
> > > - * dma_alias_devfn only supports aliases on the same bus, therefore the search
> > > + * Look for aliases to or from the given device for exisiting groups. DMA
> > 
> > s/exisiting/existing/  (Not your change, but might as well fix the
> > typo since you're touching the line anyway.)
> > 
> > > + * aliases are only supported on the same bus, therefore the search
> > >   * space is quite small (especially since we're really only looking at pcie
> > >   * device, and therefore only expect multiple slots on the root complex or
> > >   * downstream switch ports).  It's conceivable though that a pair of
> > > @@ -686,11 +686,8 @@ static struct iommu_group *get_pci_alias_group(struct pci_dev *pdev,
> > >  			continue;
> > >  
> > >  		/* We alias them or they alias us */
> > > -		if (((pdev->dev_flags & PCI_DEV_FLAGS_DMA_ALIAS_DEVFN) &&
> > > -		     pdev->dma_alias_devfn == tmp->devfn) ||
> > > -		    ((tmp->dev_flags & PCI_DEV_FLAGS_DMA_ALIAS_DEVFN) &&
> > > -		     tmp->dma_alias_devfn == pdev->devfn)) {
> > > -
> > > +		if (pci_dma_alias_is_enabled(pdev, tmp->devfn) ||
> > > +		    pci_dma_alias_is_enabled(tmp, pdev->devfn)) {
> > 
> > I was hoping for something that took two pci_dev pointers and returned
> > true if they were aliases of each other, e.g.,
> > 
> >   bool pci_dma_aliases(struct pci_dev *dev1, struct pci_dev *dev2);
> 
> Such a thing could already be created with pci_for_each_dma_alias().

Does that mean we could get replace the use of for_each_pci_dev() with some
clever use of pci_for_each_dma_alias()?