Linaro-mm-sig February 2025

linaro-mm-sig@lists.linaro.org

11 participants
25 discussions

Re: [RFC PATCH 0/5] drm/panthor: Protected mode support for Mali CSF GPUs

by Nicolas Dufresne

Le vendredi 07 février 2025 à 16:02 +0100, Boris Brezillon a écrit : > Sorry for joining the party late, a couple of comments to back Akash > and Nicolas' concerns. > > On Wed, 05 Feb 2025 13:14:14 -0500 > Nicolas Dufresne <nicolas(a)ndufresne.ca> wrote: > > > Le mercredi 05 février 2025 à 15:52 +0100, Maxime Ripard a écrit : > > > On Mon, Feb 03, 2025 at 04:43:23PM +0000, Florent Tomasin wrote: > > > > Hi Maxime, Nicolas > > > > > > > > On 30/01/2025 17:47, Nicolas Dufresne wrote: > > > > > Le jeudi 30 janvier 2025 à 17:38 +0100, Maxime Ripard a écrit : > > > > > > Hi Nicolas, > > > > > > > > > > > > On Thu, Jan 30, 2025 at 10:59:56AM -0500, Nicolas Dufresne wrote: > > > > > > > Le jeudi 30 janvier 2025 à 14:46 +0100, Maxime Ripard a écrit : > > > > > > > > Hi, > > > > > > > > > > > > > > > > I started to review it, but it's probably best to discuss it here. > > > > > > > > > > > > > > > > On Thu, Jan 30, 2025 at 01:08:56PM +0000, Florent Tomasin wrote: > > > > > > > > > Hi, > > > > > > > > > > > > > > > > > > This is a patch series covering the support for protected mode execution in > > > > > > > > > Mali Panthor CSF kernel driver. > > > > > > > > > > > > > > > > > > The Mali CSF GPUs come with the support for protected mode execution at the > > > > > > > > > HW level. This feature requires two main changes in the kernel driver: > > > > > > > > > > > > > > > > > > 1) Configure the GPU with a protected buffer. The system must provide a DMA > > > > > > > > > heap from which the driver can allocate a protected buffer. > > > > > > > > > It can be a carved-out memory or dynamically allocated protected memory region. > > > > > > > > > Some system includes a trusted FW which is in charge of the protected memory. > > > > > > > > > Since this problem is integration specific, the Mali Panthor CSF kernel > > > > > > > > > driver must import the protected memory from a device specific exporter. > > > > > > > > > > > > > > > > Why do you need a heap for it in the first place? My understanding of > > > > > > > > your series is that you have a carved out memory region somewhere, and > > > > > > > > you want to allocate from that carved out memory region your buffers. > > > > > > > > > > > > > > > > How is that any different from using a reserved-memory region, adding > > > > > > > > the reserved-memory property to the GPU device and doing all your > > > > > > > > allocation through the usual dma_alloc_* API? > > > > > > > > > > > > > > How do you then multiplex this region so it can be shared between > > > > > > > GPU/Camera/Display/Codec drivers and also userspace ? > > > > > > > > > > > > You could point all the devices to the same reserved memory region, and > > > > > > they would all allocate from there, including for their userspace-facing > > > > > > allocations. > > > > > > > > > > I get that using memory region is somewhat more of an HW description, and > > > > > aligned with what a DT is supposed to describe. One of the challenge is that > > > > > Mediatek heap proposal endup calling into their TEE, meaning knowing the region > > > > > is not that useful. You actually need the TEE APP guid and its IPC protocol. If > > > > > we can dell drivers to use a head instead, we can abstract that SoC specific > > > > > complexity. I believe each allocated addressed has to be mapped to a zone, and > > > > > that can only be done in the secure application. I can imagine similar needs > > > > > when the protection is done using some sort of a VM / hypervisor. > > > > > > > > > > Nicolas > > > > > > > > > > > > > The idea in this design is to abstract the heap management from the > > > > Panthor kernel driver (which consumes a DMA buffer from it). > > > > > > > > In a system, an integrator would have implemented a secure heap driver, > > > > and could be based on TEE or a carved-out memory with restricted access, > > > > or else. This heap driver would be responsible of implementing the > > > > logic to: allocate, free, refcount, etc. > > > > > > > > The heap would be retrieved by the Panthor kernel driver in order to > > > > allocate protected memory to load the FW and allow the GPU to enter/exit > > > > protected mode. This memory would not belong to a user space process. > > > > The driver allocates it at the time of loading the FW and initialization > > > > of the GPU HW. This is a device globally owned protected memory. > > > > > > The thing is, it's really not clear why you absolutely need to have the > > > Panthor driver involved there. It won't be transparent to userspace, > > > since you'd need an extra flag at allocation time, and the buffers > > > behave differently. If userspace has to be aware of it, what's the > > > advantage to your approach compared to just exposing a heap for those > > > secure buffers, and letting userspace allocate its buffers from there? > > > > Unless I'm mistaken, the Panthor driver loads its own firmware. Since loading > > the firmware requires placing the data in a protected memory region, and that > > this aspect has no exposure to userspace, how can Panthor not be implicated ? > > Right, the very reason we need protected memory early is because some > FW sections need to be allocated from the protected pool, otherwise the > TEE will fault as soon at the FW enters the so-called 'protected mode'. > > Now, it's not impossible to work around this limitation. For instance, > we could load the FW without this protected section by default (what we > do right now), and then provide a DRM_PANTHOR_ENABLE_FW_PROT_MODE > ioctl that would take a GEM object imported from a dmabuf allocated > from the protected dma-heap by userspace. We can then reset the FW and > allow it to operate in protected mode after that point. This approach > has two downsides though: > > 1. We have no way of checking that the memory we're passed is actually > suitable for FW execution in a protected context. If we're passed > random memory, this will likely hang the platform as soon as we enter > protected mode. > > 2. If the driver already boot the FW and exposed a DRI node, we might > have GPU workloads running, and doing a FW reset might incur a slight > delay in GPU jobs execution. > > I think #1 is a more general issue that applies to suspend buffers > allocated for GPU contexts too. If we expose ioctls where we take > protected memory buffers that can possibly lead to crashes if they are > not real protected memory regions, and we have no way to ensure the > memory is protected, we probably want to restrict these ioctls/modes to > some high-privilege CAP_SYS_. > > For #2, that's probably something we can live with, since it's a > one-shot thing. If it becomes an issue, we can even make sure we enable > the FW protected-mode before the GPU starts being used for real. > > This being said, I think the problem applies outside Panthor, and it > might be that the video codec can't reset the FW/HW block to switch to > protected mode as easily as Panthor. Overall the reset and reboot method is pretty ugly in my opinion. But to stick with the pure rationale, rebooting the SCP on MTK is much harder, since its not specific to a single HW/driver. Other codecs like Samsung MFC, Venus/Iris, Chips&Media, etc. that approach seams plausible, but we still can't trust the buffer, which to me is not acceptable. > > Note that there's also downsides to the reserved-memory node approach, > where some bootloader stage would ask the secure FW to reserve a > portion of mem and pass this through the DT. This sort of things tend to > be an integration mess, where you need all the pieces of the stack (TEE, > u-boot, MTK dma-heap driver, gbm, ...) to be at a certain version to > work properly. If we go the ioctl() way, we restrict the scope to the > TEE, gbm/mesa and the protected-dma-heap driver, which is still a lot, > but we've ripped the bootloader out of the equation at least. > > Regards, > > Boris

2 months, 2 weeks

Re: [RFC PATCH 0/5] drm/panthor: Protected mode support for Mali CSF GPUs

by Maxime Ripard

On Mon, Feb 03, 2025 at 04:43:23PM +0000, Florent Tomasin wrote: > Hi Maxime, Nicolas > > On 30/01/2025 17:47, Nicolas Dufresne wrote: > > Le jeudi 30 janvier 2025 à 17:38 +0100, Maxime Ripard a écrit : > >> Hi Nicolas, > >> > >> On Thu, Jan 30, 2025 at 10:59:56AM -0500, Nicolas Dufresne wrote: > >>> Le jeudi 30 janvier 2025 à 14:46 +0100, Maxime Ripard a écrit : > >>>> Hi, > >>>> > >>>> I started to review it, but it's probably best to discuss it here. > >>>> > >>>> On Thu, Jan 30, 2025 at 01:08:56PM +0000, Florent Tomasin wrote: > >>>>> Hi, > >>>>> > >>>>> This is a patch series covering the support for protected mode execution in > >>>>> Mali Panthor CSF kernel driver. > >>>>> > >>>>> The Mali CSF GPUs come with the support for protected mode execution at the > >>>>> HW level. This feature requires two main changes in the kernel driver: > >>>>> > >>>>> 1) Configure the GPU with a protected buffer. The system must provide a DMA > >>>>> heap from which the driver can allocate a protected buffer. > >>>>> It can be a carved-out memory or dynamically allocated protected memory region. > >>>>> Some system includes a trusted FW which is in charge of the protected memory. > >>>>> Since this problem is integration specific, the Mali Panthor CSF kernel > >>>>> driver must import the protected memory from a device specific exporter. > >>>> > >>>> Why do you need a heap for it in the first place? My understanding of > >>>> your series is that you have a carved out memory region somewhere, and > >>>> you want to allocate from that carved out memory region your buffers. > >>>> > >>>> How is that any different from using a reserved-memory region, adding > >>>> the reserved-memory property to the GPU device and doing all your > >>>> allocation through the usual dma_alloc_* API? > >>> > >>> How do you then multiplex this region so it can be shared between > >>> GPU/Camera/Display/Codec drivers and also userspace ? > >> > >> You could point all the devices to the same reserved memory region, and > >> they would all allocate from there, including for their userspace-facing > >> allocations. > > > > I get that using memory region is somewhat more of an HW description, and > > aligned with what a DT is supposed to describe. One of the challenge is that > > Mediatek heap proposal endup calling into their TEE, meaning knowing the region > > is not that useful. You actually need the TEE APP guid and its IPC protocol. If > > we can dell drivers to use a head instead, we can abstract that SoC specific > > complexity. I believe each allocated addressed has to be mapped to a zone, and > > that can only be done in the secure application. I can imagine similar needs > > when the protection is done using some sort of a VM / hypervisor. > > > > Nicolas > > > > The idea in this design is to abstract the heap management from the > Panthor kernel driver (which consumes a DMA buffer from it). > > In a system, an integrator would have implemented a secure heap driver, > and could be based on TEE or a carved-out memory with restricted access, > or else. This heap driver would be responsible of implementing the > logic to: allocate, free, refcount, etc. > > The heap would be retrieved by the Panthor kernel driver in order to > allocate protected memory to load the FW and allow the GPU to enter/exit > protected mode. This memory would not belong to a user space process. > The driver allocates it at the time of loading the FW and initialization > of the GPU HW. This is a device globally owned protected memory. The thing is, it's really not clear why you absolutely need to have the Panthor driver involved there. It won't be transparent to userspace, since you'd need an extra flag at allocation time, and the buffers behave differently. If userspace has to be aware of it, what's the advantage to your approach compared to just exposing a heap for those secure buffers, and letting userspace allocate its buffers from there? > When I came across this patch series: > - > https://lore.kernel.org/lkml/20230911023038.30649-1-yong.wu@mediatek.com/#t > I found it could help abstract the interface between the secure heap and > the integration of protected memory in Panthor. > > A kernel driver would have to find the heap: `dma_heap_find()`, then > request allocation of a DMA buffer from it. The heap driver would deal > with the specifities of the protected memory on the system. Sure, but we still have to address *why* it would be a good idea for the driver to do it in the first place. The mediatek series had the same feedback. Maxime

2 months, 2 weeks

Re: [RFC PATCH 0/5] drm/panthor: Protected mode support for Mali CSF GPUs

by Nicolas Dufresne

Le lundi 03 février 2025 à 16:43 +0000, Florent Tomasin a écrit : > Hi Maxime, Nicolas > > On 30/01/2025 17:47, Nicolas Dufresne wrote: > > Le jeudi 30 janvier 2025 à 17:38 +0100, Maxime Ripard a écrit : > > > Hi Nicolas, > > > > > > On Thu, Jan 30, 2025 at 10:59:56AM -0500, Nicolas Dufresne wrote: > > > > Le jeudi 30 janvier 2025 à 14:46 +0100, Maxime Ripard a écrit : > > > > > Hi, > > > > > > > > > > I started to review it, but it's probably best to discuss it here. > > > > > > > > > > On Thu, Jan 30, 2025 at 01:08:56PM +0000, Florent Tomasin wrote: > > > > > > Hi, > > > > > > > > > > > > This is a patch series covering the support for protected mode execution in > > > > > > Mali Panthor CSF kernel driver. > > > > > > > > > > > > The Mali CSF GPUs come with the support for protected mode execution at the > > > > > > HW level. This feature requires two main changes in the kernel driver: > > > > > > > > > > > > 1) Configure the GPU with a protected buffer. The system must provide a DMA > > > > > > heap from which the driver can allocate a protected buffer. > > > > > > It can be a carved-out memory or dynamically allocated protected memory region. > > > > > > Some system includes a trusted FW which is in charge of the protected memory. > > > > > > Since this problem is integration specific, the Mali Panthor CSF kernel > > > > > > driver must import the protected memory from a device specific exporter. > > > > > > > > > > Why do you need a heap for it in the first place? My understanding of > > > > > your series is that you have a carved out memory region somewhere, and > > > > > you want to allocate from that carved out memory region your buffers. > > > > > > > > > > How is that any different from using a reserved-memory region, adding > > > > > the reserved-memory property to the GPU device and doing all your > > > > > allocation through the usual dma_alloc_* API? > > > > > > > > How do you then multiplex this region so it can be shared between > > > > GPU/Camera/Display/Codec drivers and also userspace ? > > > > > > You could point all the devices to the same reserved memory region, and > > > they would all allocate from there, including for their userspace-facing > > > allocations. > > > > I get that using memory region is somewhat more of an HW description, and > > aligned with what a DT is supposed to describe. One of the challenge is that > > Mediatek heap proposal endup calling into their TEE, meaning knowing the region > > is not that useful. You actually need the TEE APP guid and its IPC protocol. If > > we can dell drivers to use a head instead, we can abstract that SoC specific > > complexity. I believe each allocated addressed has to be mapped to a zone, and > > that can only be done in the secure application. I can imagine similar needs > > when the protection is done using some sort of a VM / hypervisor. > > > > Nicolas > > > > The idea in this design is to abstract the heap management from the > Panthor kernel driver (which consumes a DMA buffer from it). > > In a system, an integrator would have implemented a secure heap driver, > and could be based on TEE or a carved-out memory with restricted access, > or else. This heap driver would be responsible of implementing the > logic to: allocate, free, refcount, etc. > > The heap would be retrieved by the Panthor kernel driver in order to > allocate protected memory to load the FW and allow the GPU to enter/exit > protected mode. This memory would not belong to a user space process. > The driver allocates it at the time of loading the FW and initialization > of the GPU HW. This is a device globally owned protected memory. This use case also applies well for codec. The Mediatek SCP firmware needs to be loaded with a restricted memory too, its a very similar scenario, plus Mediatek chips often include a Mali. On top of that, V4L2 codecs (in general) do need to allocate internal scratch buffer for the IP to write to for things like motion vectors, reconstruction frames, entropy statistics, etc. The IP will only be able to write if the memory is restricted. Nicolas > > When I came across this patch series: > - > https://lore.kernel.org/lkml/20230911023038.30649-1-yong.wu@mediatek.com/#t > I found it could help abstract the interface between the secure heap and > the integration of protected memory in Panthor. > > A kernel driver would have to find the heap: `dma_heap_find()`, then > request allocation of a DMA buffer from it. The heap driver would deal > with the specifities of the protected memory on the system. > > > > > > > > Also, how the secure memory is allocted / obtained is a process that > > > > can vary a lot between SoC, so implementation details assumption > > > > should not be coded in the driver. > > > > > > But yeah, we agree there, it's also the point I was trying to make :) > > > > > > Maxime > > > > Agree with your point, the Panthor kernel driver may not be aware of the > heap management logic. As an alternative to the DMA heap API used here, > I also tried to expose the heap by passing the phandle of a "heap" > device to Panthor. The reference to the DMA heap was stored as a private > data of the heap device as a new type: `struct dma_heap_import_info`, > similar to the existing type: `struct dma_heap_export_info`. > This made me think it could be problematic, as the private data type > would have to be cast before accessing it from the importer driver. I > worried about a mis-use of the types with this approach. > > Regards, > Florent

2 months, 2 weeks

❌ Incoming Messages Blocked

by lists.linaro.org

2 months, 2 weeks

Re: [RFC PATCH 1/5] dt-bindings: dma: Add CMA Heap bindings

by Nicolas Dufresne

Hi Florent, Le lundi 03 février 2025 à 13:36 +0000, Florent Tomasin a écrit : > > On 30/01/2025 13:28, Maxime Ripard wrote: > > Hi, > > > > On Thu, Jan 30, 2025 at 01:08:57PM +0000, Florent Tomasin wrote: > > > Introduce a CMA Heap dt-binding allowing custom > > > CMA heap registrations. > > > > > > * Note to the reviewers: > > > The patch was used for the development of the protected mode Just to avoid divergence in nomenclature, and because this is not a new subject, perhaps you should also adhere to the name "restricted". Both Linaro and Mediatek have moved from "secure" to that name in their proposal. As you are the third proposing this (at least for the proposal that are CCed on linux-media), I would have expected in your cover letter a summary of how the other requirement have been blended in your proposal. regards, Nicolas > > > feature in Panthor CSF kernel driver and is not initially thought > > > to land in the Linux kernel. It is mostly relevant if someone > > > wants to reproduce the environment of testing. Please, raise > > > interest if you think the patch has value in the Linux kernel. > > > > > > Signed-off-by: Florent Tomasin <florent.tomasin(a)arm.com> > > > --- > > > .../devicetree/bindings/dma/linux,cma.yml | 43 +++++++++++++++++++ > > > 1 file changed, 43 insertions(+) > > > create mode 100644 Documentation/devicetree/bindings/dma/linux,cma.yml > > > > > > diff --git a/Documentation/devicetree/bindings/dma/linux,cma.yml b/Documentation/devicetree/bindings/dma/linux,cma.yml > > > new file mode 100644 > > > index 000000000000..c532e016bbe5 > > > --- /dev/null > > > +++ b/Documentation/devicetree/bindings/dma/linux,cma.yml > > > @@ -0,0 +1,43 @@ > > > +# SPDX-License-Identifier: GPL-2.0-only OR BSD-2-Clause > > > +%YAML 1.2 > > > +--- > > > +$id: http://devicetree.org/schemas/dma/linux-cma.yaml# > > > +$schema: http://devicetree.org/meta-schemas/core.yaml# > > > + > > > +title: Custom Linux CMA heap > > > + > > > +description: > > > + The custom Linux CMA heap device tree node allows registering > > > + of multiple CMA heaps. > > > + > > > + The CMA heap name will match the node name of the "memory-region". > > > + > > > +properties: > > > + compatible: > > > + enum: > > > + - linux,cma > > > + > > > + memory-region: > > > + maxItems: 1 > > > + description: | > > > + Phandle to the reserved memory node associated with the CMA Heap. > > > + The reserved memory node must follow this binding convention: > > > + - Documentation/devicetree/bindings/reserved-memory/reserved-memory.txt > > > + > > > +examples: > > > + - | > > > + reserved-memory { > > > + #address-cells = <2>; > > > + #size-cells = <2>; > > > + > > > + custom_cma_heap: custom-cma-heap { > > > + compatible = "shared-dma-pool"; > > > + reg = <0x0 0x90600000 0x0 0x1000000>; > > > + reusable; > > > + }; > > > + }; > > > + > > > + device_cma_heap: device-cma-heap { > > > + compatible = "linux,cma"; > > > + memory-region = <&custom_cma_heap>; > > > + }; > > > > Isn't it redundant with the linux,cma-default shared-dma-pool property > > already? > > > > Maxime > > Hi Maxime, > > Please correct me if my understanding is wrong, > > The existing properties: linux,cma-default and shared-dma-pool, do not > allow the creations of multiple standalone CMA heaps, those will create > a single CMA heap: `dma_contiguous_default_area`? Other CMA heaps will > be bound to a driver. > > I introduced the "linux,cma" to allow creating multiple standalone CMA > heaps, with the intention of validating the protected mode support on > Mali CSG GPUs. It was included in the RFC in there are interests in > this approach. > > Since the Panthor CSF kernel driver does not own or manage a heap, > I needed a way to create a standalone heap. The idea here is for the > kernel driver to be an importer. I relied on a patch series to retrieve > the heap and allocate a DMA buffer from it: > - dma_heap_find() > - dma_heap_buffer_alloc() > - dma_heap_put() > > Ref: > https://lore.kernel.org/lkml/20230911023038.30649-1-yong.wu@mediatek.com/#t > > > Since the protected/secure memory management is integration specific, > I needed a generic way for Panthor to allocate from such heap. > > In some scenarios it might be a carved-out memory, in others a FW will > reside in the system (TEE) and require a secure heap driver to allocate > memory (e.g: a similar approach is followd by MTK). Such driver would > implement the allocation and free logic. > > Florent > > >

2 months, 2 weeks

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

Linaro-mm-sig February 2025