On Tue, Jul 15, 2014 at 9:25 PM, Olav Haugan [off-list ref] wrote:

On 7/13/2014 4:43 AM, Rob Clark wrote:

quoted

On Sun, Jul 13, 2014 at 5:43 AM, Will Deacon [off-list ref] wrote:

quoted

On Sat, Jul 12, 2014 at 01:57:31PM +0100, Rob Clark wrote:

quoted

On Sat, Jul 12, 2014 at 8:22 AM, Arnd Bergmann [off-list ref] wrote:

quoted

On Saturday 12 July 2014, Rob Clark wrote:

quoted

quoted

quoted

Was there actually a good reason for having the device link to the
iommu rather than the other way around?  How much would people hate it
if I just ignore the generic bindings and use something that works for
me instead.  I mean, it isn't exactly like there is going to be .dts
re-use across different SoC's..  and at least with current IOMMU API
some sort of of_get_named_iommu() API doesn't really make sense.

The thing is, if you end up ignoring the generic binding then we have two
IOMMUs using the same (ARM SMMU) binding and it begs the question as to
which is the more generic! I know we're keen to get this merged, but merging
something that people won't use and calling it generic doesn't seem ideal
either. We do, however, desperately need a generic binding.

yeah, ignoring the generic binding is not my first choice.  I'd rather
have something that works well for everyone.  But I wasn't really sure
if the current proposal was arbitrary, or if there are some
conflicting requirements between different platforms.

The common case that needs to be simple is attaching one (master) device
to an IOMMU using the shared global context for the purposes of implementing
the dma-mapping API.

well, I don't disagree that IOMMU API has some problems.  It is too
tied to the bus type, which doesn't really seem to make sense for
platform devices.  (Unless we start having multiple platform busses?)

But at least given the current IOMMU API I'm not really sure how it
makes a difference which way the link goes.  But if there has already
been some discussion about how you want to handle the tie in with
dma-mapping, if you could point me at that then maybe your point will
make more sense to me.

If you look at the proposed binding in isolation, I think it *is* cleaner
than the ARM SMMU binding (I've acked it...) and I believe it's more
consistent with the way we describe linkages elsewhere.

However, the issue you're raising is that it's more difficult to make use of
in a Linux IOMMU driver. The reward you'll get for using it will come
eventually when the DMA ops are automatically swizzled for devices using the
generic binding.

My plan for the ARM SMMU driver is:

  (1) Change ->probe() to walk the device-tree looking for all masters with
      phandles back to the SMMU instance being probed

  (2) For each master, extract the Stream IDs and add them to the internal
      SMMU driver data structures (an rbtree per SMMU instance). For
      hotpluggable buses, we'll need a way for the bus controller to
      reserve a range of IDs -- this will likely be a later extension to
      the binding.

  (3) When we get an ->add() call, warn if it's a device we haven't seen
      and reject the addition.

That way, ->attach() should be the same as it is now, I think.

Have you tried implementing something like that? We agreed that (1) isn't
pretty, but I don't have a good alternative and it's only done at
probe-time.

I haven't tried implementing that yet, but I'm sure it would work.  I
was just hoping to avoid having to do that ;-)

Is the reason you want to do it this way because you want to guarantee
that all masters (and stream IDs) have been identified before the first
attach call? I am just wondering why you cannot continue doing the
master/streamID discovery during add_device() callback?

it was mostly because I couldn't think of a sane way to differentiate
between first and second time a device attaches (without keeping a
reference to the device).  But I guess it is ok to assume no hotplug
(since walking the device tree also seems acceptable)

BR,
-R

quoted

quoted

BTW: Is the msm-v0 IOMMU compatible with the ARM SMMU driver, or is it a
completely different design requiring a different driver?

My understanding is that it is different from msm v1 IOMMU, although I
think it shares the same pagetable format with v1.  Not sure if that
is the same as arm-smmu?   If so it might be nice to try to extract
out some shared helper fxns for map/unmap as well.

I expect Olav knows better the similarities/differences.

The msm-v0 IOMMU is not compatible with ARM SMMUv1 specification.
However, it is a close cousin. The hardware was designed before the ARM
SMMUv1 specification was available I believe. But it shares many of the
same concepts as the ARM SMMUv1.

msm-v0 IOMMU supports V7S page table format only. The ARM SMMU driver
does not support V7S at this time. However, I believe we need to support
this.

Will, this reminds me. We definitely have a need to use different page
tables in the ARM SMMU driver vs. the ARM CPU. We have an SoC with ARMv8
cores (and thus ARMv8 page tables) but the SMMUs (SMMUv1) on this SoC
only have support for V7S/V7L page table format. So we cannot use the
same page table format as the CPU.

Thanks,

Olav

--
The Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
hosted by The Linux Foundation