Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
From:   Tom Murphy <murphyt7@tcd.ie>
To:     iommu@lists.linux-foundation.org
Cc:     Tom Murphy <murphyt7@tcd.ie>, Joerg Roedel <joro@8bytes.org>,
        David Woodhouse <dwmw2@infradead.org>,
        Lu Baolu <baolu.lu@linux.intel.com>,
        linux-kernel@vger.kernel.org
Subject: [PATCH V2 0/5] Convert the intel iommu driver to the dma-iommu api
Date:   Thu,  3 Sep 2020 21:18:32 +0100
Message-Id: <20200903201839.7327-1-murphyt7@tcd.ie>
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

This patchset converts the intel iommu driver to the dma-iommu api.

While converting the driver I exposed a bug in the intel i915 driver which causes a huge amount of artifacts on the screen of my laptop. You can see a picture of it here:
https://github.com/pippy360/kernelPatches/blob/master/IMG_20191219_225922.jpg

This issue is in the i915 driver and is caused by the driver not respecting the return value of the dma_map_ops::map_sg function.

We talked about this in this microconference:
https://linuxplumbersconf.org/event/7/contributions/846/
and came to the conclusion that we should add an attribute to disable combining sg segments in the dma-iommu api (in the __finalise_sg function). This will work as a temporary fix and allow us to convert the intel iommu driver to the dma-iommu path while we wait for the i915 driver to be rewritten to respect the return value of map_sg. I haven't done this work yet and won't have time to do it. If someone else could take this on that would be great.

To allow my patch set to be tested I have added a patch (the "DO NOT MERGE..." patch) in this series to disable combining sg segments in the dma-iommu api which fixes the bug but it doesn't fix the actual problem.

As part of this patch series I copied the intel bounce buffer code to the dma-iommu path. The addition of the bounce buffer code took me by surprise. I did most of my development on this patch series before the bounce buffer code was added and my reimplementation in the dma-iommu path is very rushed and not properly tested but I’m running out of time to work on this patch set.

On top of that I also didn’t port over the intel tracing code from this commit:
https://github.com/torvalds/linux/commit/3b53034c268d550d9e8522e613a14ab53b8840d8#diff-6b3e7c4993f05e76331e463ab1fc87e1
So all the work in that commit is now wasted. The code will need to be removed and reimplemented in the dma-iommu path. I would like to take the time to do this but I really don’t have the time at the moment and I want to get these changes out before the iommu code changes any more.

Unfortunately I no longer have enough spare time to continue to work on/rebase this patch series. So this will most likely be the last patch series from me for the intel dma-iommu conversion.

Change-log:
v2:
-Rebase on top of the latest staging branch
-move the freelist parameter to iommu_iotlb_gather

Signed-off-by: Tom Murphy <murphyt7@tcd.ie>

Tom Murphy (5):
  iommu: Handle freelists when using deferred flushing in iommu drivers
  iommu: Add iommu_dma_free_cpu_cached_iovas function
  iommu: allow the dma-iommu api to use bounce buffers
  iommu/vt-d: Convert intel iommu driver to the iommu ops
  DO NOT MERGE: iommu: disable list appending in dma-iommu

 drivers/iommu/Kconfig       |   1 +
 drivers/iommu/dma-iommu.c   | 169 +++++---
 drivers/iommu/intel/iommu.c | 805 ++++--------------------------------
 drivers/iommu/iommu.c       |  10 +
 include/linux/dma-iommu.h   |   3 +
 include/linux/iommu.h       |   8 +
 6 files changed, 222 insertions(+), 774 deletions(-)

-- 
2.20.1