lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <c382fb0e-6b73-5ca0-7f63-d2843351325e@nvidia.com>
Date:   Mon, 6 Dec 2021 09:55:09 +0000
From:   Jon Hunter <jonathanh@...dia.com>
To:     joro@...tes.org, will@...nel.org, robh+dt@...nel.org,
        robin.murphy@....com
Cc:     linux-tegra@...r.kernel.org, dri-devel@...ts.freedesktop.org,
        iommu@...ts.linux-foundation.org, linux-kernel@...r.kernel.org,
        devicetree@...r.kernel.org, linux-arm-kernel@...ts.infradead.org,
        Mikko Perttunen <cyndis@...si.fi>,
        Mikko Perttunen <mperttunen@...dia.com>,
        thierry.reding@...il.com
Subject: Re: [PATCH v2 0/8] Host1x context isolation support

Will, Joerg, Rob,

On 08/11/2021 10:36, Mikko Perttunen wrote:
> On 9/16/21 5:32 PM, Mikko Perttunen wrote:
>> Hi all,
>>
>> ***
>> New in v2:
>>
>> Added support for Tegra194
>> Use standard iommu-map property instead of custom mechanism
>> ***
>>
>> this series adds support for Host1x 'context isolation'. Since
>> when programming engines through Host1x, userspace can program in
>> any addresses it wants, we need some way to isolate the engines'
>> memory spaces. Traditionally this has either been done imperfectly
>> with a single shared IOMMU domain, or by copying and verifying the
>> programming command stream at submit time (Host1x firewall).
>>
>> Since Tegra186 there is a privileged (only usable by kernel)
>> Host1x opcode that allows setting the stream ID sent by the engine
>> to the SMMU. So, by allocating a number of context banks and stream
>> IDs for this purpose, and using this opcode at the beginning of
>> each job, we can implement isolation. Due to the limited number of
>> context banks only each process gets its own context, and not
>> each channel.
>>
>> This feature also allows sharing engines among multiple VMs when
>> used with Host1x's hardware virtualization support - up to 8 VMs
>> can be configured with a subset of allowed stream IDs, enforced
>> at hardware level.
>>
>> To implement this, this series adds a new host1x context bus, which
>> will contain the 'struct device's corresponding to each context
>> bank / stream ID, changes to device tree and SMMU code to allow
>> registering the devices and using the bus, as well as the Host1x
>> stream ID programming code and support in TegraDRM.
>>
>> Device tree bindings are not updated yet pending consensus that the
>> proposed changes make sense.
>>
>> Thanks,
>> Mikko
>>
>> Mikko Perttunen (8):
>>    gpu: host1x: Add context bus
>>    gpu: host1x: Add context device management code
>>    gpu: host1x: Program context stream ID on submission
>>    iommu/arm-smmu: Attach to host1x context device bus
>>    arm64: tegra: Add Host1x context stream IDs on Tegra186+
>>    drm/tegra: falcon: Set DMACTX field on DMA transactions
>>    drm/tegra: vic: Implement get_streamid_offset
>>    drm/tegra: Support context isolation
>>
>>   arch/arm64/boot/dts/nvidia/tegra186.dtsi  |  12 ++
>>   arch/arm64/boot/dts/nvidia/tegra194.dtsi  |  12 ++
>>   drivers/gpu/Makefile                      |   3 +-
>>   drivers/gpu/drm/tegra/drm.h               |   2 +
>>   drivers/gpu/drm/tegra/falcon.c            |   8 +
>>   drivers/gpu/drm/tegra/falcon.h            |   1 +
>>   drivers/gpu/drm/tegra/submit.c            |  13 ++
>>   drivers/gpu/drm/tegra/uapi.c              |  34 ++++-
>>   drivers/gpu/drm/tegra/vic.c               |  38 +++++
>>   drivers/gpu/host1x/Kconfig                |   5 +
>>   drivers/gpu/host1x/Makefile               |   2 +
>>   drivers/gpu/host1x/context.c              | 174 ++++++++++++++++++++++
>>   drivers/gpu/host1x/context.h              |  27 ++++
>>   drivers/gpu/host1x/context_bus.c          |  31 ++++
>>   drivers/gpu/host1x/dev.c                  |  12 +-
>>   drivers/gpu/host1x/dev.h                  |   2 +
>>   drivers/gpu/host1x/hw/channel_hw.c        |  52 ++++++-
>>   drivers/gpu/host1x/hw/host1x06_hardware.h |  10 ++
>>   drivers/gpu/host1x/hw/host1x07_hardware.h |  10 ++
>>   drivers/iommu/arm/arm-smmu/arm-smmu.c     |  13 ++
>>   include/linux/host1x.h                    |  21 +++
>>   include/linux/host1x_context_bus.h        |  15 ++
>>   22 files changed, 488 insertions(+), 9 deletions(-)
>>   create mode 100644 drivers/gpu/host1x/context.c
>>   create mode 100644 drivers/gpu/host1x/context.h
>>   create mode 100644 drivers/gpu/host1x/context_bus.c
>>   create mode 100644 include/linux/host1x_context_bus.h
>>
> 
> IOMMU/DT folks, any thoughts about this approach? The patches that are 
> of interest outside of Host1x/TegraDRM specifics are patches 1, 2, 4, 
> and 5.


Any feedback on this?

Jon

-- 
nvpublic

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ