[00/18] Convert arch/arm to use iommu-dma

Hi all,

After 5 years or so of intending to get round to this, finally the
time comes! The changes themselves actualy turn out to be relatively
mechanical; the bigger concern appears to be how to get everything
merged across about 5 diffferent trees given the dependencies.

I've lightly boot-tested things on Rockchip RK3288 and Exynos 4412
(Odroid-U3), to the degree that their display drivers should be using
IOMMU-backed buffers and don't explode (the Odroid doesn't manage to
send a working HDMI signal to the one monitor I have that it actually
detects, but that's a pre-existing condition...) Confirmation that the
Mediatek, OMAP and Tegra changes work will be most welcome.

Patches are based on 5.9-rc1, branch available here:

  git://linux-arm.org/linux-rm arm/dma


Robin.


Robin Murphy (18):
  ARM/dma-mapping: Drop .dma_supported for IOMMU ops
  ARM/dma-mapping: Consolidate IOMMU ops callbacks
  ARM/dma-mapping: Merge IOMMU ops
  iommu/dma: Add temporary hacks for arch/arm
  ARM/dma-mapping: Switch to iommu_dma_ops
  ARM/dma-mapping: Support IOMMU default domains
  iommu/arm-smmu: Remove arch/arm workaround
  iommu/renesas: Remove arch/arm workaround
  iommu/mediatek-v1: Add IOMMU_DOMAIN_DMA support
  iommu/msm: Add IOMMU_DOMAIN_DMA support
  iommu/omap: Add IOMMU_DOMAIN_DMA support
  iommu/tegra-gart: Add IOMMU_DOMAIN_DMA support
  iommu/tegra: Add IOMMU_DOMAIN_DMA support
  drm/exynos: Consolidate IOMMU mapping code
  drm/nouveau/tegra: Clean up IOMMU workaround
  staging/media/tegra-vde: Clean up IOMMU workaround
  media/omap3isp: Clean up IOMMU workaround
  ARM/dma-mapping: Remove legacy dma-iommu API

 arch/arm/Kconfig                              |   28 +-
 arch/arm/common/dmabounce.c                   |    1 -
 arch/arm/include/asm/device.h                 |    9 -
 arch/arm/include/asm/dma-iommu.h              |   37 -
 arch/arm/mm/dma-mapping.c                     | 1198 +----------------
 drivers/gpu/drm/exynos/exynos5433_drm_decon.c |    5 +-
 drivers/gpu/drm/exynos/exynos7_drm_decon.c    |    5 +-
 drivers/gpu/drm/exynos/exynos_drm_dma.c       |   61 +-
 drivers/gpu/drm/exynos/exynos_drm_drv.h       |    6 +-
 drivers/gpu/drm/exynos/exynos_drm_fimc.c      |    5 +-
 drivers/gpu/drm/exynos/exynos_drm_fimd.c      |    5 +-
 drivers/gpu/drm/exynos/exynos_drm_g2d.c       |    5 +-
 drivers/gpu/drm/exynos/exynos_drm_gsc.c       |    5 +-
 drivers/gpu/drm/exynos/exynos_drm_rotator.c   |    5 +-
 drivers/gpu/drm/exynos/exynos_drm_scaler.c    |    6 +-
 drivers/gpu/drm/exynos/exynos_mixer.c         |    7 +-
 .../drm/nouveau/nvkm/engine/device/tegra.c    |   13 -
 drivers/iommu/Kconfig                         |    8 -
 drivers/iommu/arm/arm-smmu/arm-smmu.c         |   10 -
 drivers/iommu/ipmmu-vmsa.c                    |   69 -
 drivers/iommu/msm_iommu.c                     |    7 +-
 drivers/iommu/mtk_iommu.h                     |    2 -
 drivers/iommu/mtk_iommu_v1.c                  |  153 +--
 drivers/iommu/omap-iommu.c                    |   22 +-
 drivers/iommu/tegra-gart.c                    |   17 +-
 drivers/iommu/tegra-smmu.c                    |   37 +-
 drivers/media/platform/Kconfig                |    1 -
 drivers/media/platform/omap3isp/isp.c         |   68 +-
 drivers/media/platform/omap3isp/isp.h         |    3 -
 drivers/staging/media/tegra-vde/iommu.c       |   12 -
 30 files changed, 150 insertions(+), 1660 deletions(-)
 delete mode 100644 arch/arm/include/asm/dma-iommu.h
  

Hi Robin,

On 20.08.2020 17:08, Robin Murphy wrote:
> Hi all,
>
> After 5 years or so of intending to get round to this, finally the
> time comes! The changes themselves actualy turn out to be relatively
> mechanical; the bigger concern appears to be how to get everything
> merged across about 5 diffferent trees given the dependencies.
>
> I've lightly boot-tested things on Rockchip RK3288 and Exynos 4412
> (Odroid-U3), to the degree that their display drivers should be using
> IOMMU-backed buffers and don't explode (the Odroid doesn't manage to
> send a working HDMI signal to the one monitor I have that it actually
> detects, but that's a pre-existing condition...) Confirmation that the
> Mediatek, OMAP and Tegra changes work will be most welcome.
>
> Patches are based on 5.9-rc1, branch available here:
>
>    git://linux-arm.org/linux-rm arm/dma

Well, my first proposal for the ARM and ARM64 DMA-mapping unification 
has been posted long time ago: https://lkml.org/lkml/2016/2/19/79

Thanks for resurrecting it! :)

I've tested this patchset on various ARM32bit Exynos based boards (not 
only Exynos4412) and most of them works fine after your conversion. 
However there are issues you cannot learn from the code.

Conversion of the Exynos DRM was straightforward (thanks!), but there 
are other Exynos drivers that depends on the old ARM implementation. The 
S5P-MFC (only for the v5 hardware) and Exynos4 FIMC-ISP drivers depends 
on the first-fit IOVA allocation algorithm in the old ARM DMA-mapping. 
This was the main reason I've didn't continue my initial conversion attempt.

Both drivers allocate a buffer for their firmware and then in the 
hardware registers address video buffers as an offset from the 
begginning of the firmware. This doesn't work when underlying 
DMA-mapping allocates IOVA with the last-fit algorithm, what the 
drivers/iommu/dma-iommu.c does. So far I didn't find a good solution for 
that issue.

I'm open for suggestions. One more limitation for the S5P-MFC driver is 
that the hardware is capable only for addressing 128MiB. They will 
probably need to call IOMMU API directly, but I would like to keep as 
much from the IOMMU/DMA-mapping code as possible.


Anyway, we need to move ARM 32bit forward, so for the ARM DMA-mapping 
and Exynos DRM changes, feel free to add:

Acked-by: Marek Szyprowski <m.szyprowski@samsung.com>

Tested-by: Marek Szyprowski <m.szyprowski@samsung.com>

Best regards
  

Tested full series on bananapi r2 (mt7623/mt2701, 5.9-rc1 + hdmi-patches), works so far fbcon+x without issues

Tested-by: Frank Wunderlich <frank-w@public-files.de>

regards Frank
  

On 27/08/2020 14:31, Frank Wunderlich wrote:
> Tested full series on bananapi r2 (mt7623/mt2701, 5.9-rc1 + hdmi-patches), works so far fbcon+x without issues
> 
> Tested-by: Frank Wunderlich <frank-w@public-files.de>
> 

Thanks for testing.

Robin this is especially relevant for:
[PATCH 09/18] iommu/mediatek-v1: Add IOMMU_DOMAIN_DMA support

Regards,
Matthias
  

Hi all,

On 24.08.2020 13:40, Marek Szyprowski wrote:
> On 20.08.2020 17:08, Robin Murphy wrote:
>> After 5 years or so of intending to get round to this, finally the
>> time comes! The changes themselves actualy turn out to be relatively
>> mechanical; the bigger concern appears to be how to get everything
>> merged across about 5 diffferent trees given the dependencies.
>>
>> I've lightly boot-tested things on Rockchip RK3288 and Exynos 4412
>> (Odroid-U3), to the degree that their display drivers should be using
>> IOMMU-backed buffers and don't explode (the Odroid doesn't manage to
>> send a working HDMI signal to the one monitor I have that it actually
>> detects, but that's a pre-existing condition...) Confirmation that the
>> Mediatek, OMAP and Tegra changes work will be most welcome.
>>
>> Patches are based on 5.9-rc1, branch available here:
>>
>>    git://linux-arm.org/linux-rm arm/dma
>
> Well, my first proposal for the ARM and ARM64 DMA-mapping unification 
> has been posted long time ago: https://lkml.org/lkml/2016/2/19/79
>
> Thanks for resurrecting it! :)
>
> I've tested this patchset on various ARM32bit Exynos based boards (not 
> only Exynos4412) and most of them works fine after your conversion. 
> However there are issues you cannot learn from the code.
>
> Conversion of the Exynos DRM was straightforward (thanks!), but there 
> are other Exynos drivers that depends on the old ARM implementation. 
> The S5P-MFC (only for the v5 hardware) and Exynos4 FIMC-ISP drivers 
> depends on the first-fit IOVA allocation algorithm in the old ARM 
> DMA-mapping. This was the main reason I've didn't continue my initial 
> conversion attempt.
>
> Both drivers allocate a buffer for their firmware and then in the 
> hardware registers address video buffers as an offset from the 
> begginning of the firmware. This doesn't work when underlying 
> DMA-mapping allocates IOVA with the last-fit algorithm, what the 
> drivers/iommu/dma-iommu.c does. So far I didn't find a good solution 
> for that issue.
>
> I'm open for suggestions. One more limitation for the S5P-MFC driver 
> is that the hardware is capable only for addressing 128MiB. They will 
> probably need to call IOMMU API directly, but I would like to keep as 
> much from the IOMMU/DMA-mapping code as possible.

Just for the record. I've finally managed to add needed workarounds to 
the both problematic Exynos4 drivers, so they work fine with this 
patchset. It turned out that it wasn't that hard:

https://lore.kernel.org/linux-samsung-soc/20200918144833.14618-1-m.szyprowski@samsung.com/T/#t

So from my side you have a green light to go ahead and switch ARM 32bit 
to generic code. Time to say good bye to the one of my biggest 
architecture related things merged once to mainline Linux. ;)

Best regards