| Message ID | 20221117185424.2359687-1-thierry.reding@gmail.com |
|---|---|
| Headers | show |
| Series | iommu: Support mappings/reservations in reserved-memory regions | expand |
On Thu, Nov 17, 2022 at 07:54:20PM +0100, Thierry Reding wrote: > From: Thierry Reding <treding@nvidia.com> > > Hi, > > This version is a minor update to the previous v11, which can be found > here: > > https://lore.kernel.org/all/20221111161806.630527-1-thierry.reding@gmail.com/ > > The only change here is that the #dma-{address,size}-cells is dropped. > It turns out to be much simpler to just update #{address,size}-cells to > what they should be rather than add extra complexity for the DMA work- > around. There's a minor update to the DT binding so that it can now > properly validate cases where we have both reg and iommu-addresses > properties. > > An example is included in the DT bindings, but here is an extract of > what I've used to test this: > > reserved-memory { > #address-cells = <2>; > #size-cells = <2>; > ranges; > > /* > * Creates an identity mapping for the framebuffer that > * the firmware has setup to scan out a bootsplash from. > */ > fb: framebuffer@92cb2000 { > reg = <0x0 0x92cb2000 0x0 0x00800000>; > iommu-addresses = <&dc0 0x0 0x92cb2000 0x0 0x00800000>; > }; > > /* > * Creates a reservation in the IOVA space to prevent > * any buffers from being mapped to that region. Note > * that on Tegra the range is actually quite different > * from this, but it would conflict with the display > * driver that I tested this against, so this is just > * a dummy region for testing. > */ > adsp: reservation-adsp { > iommu-addresses = <&dc0 0x0 0x90000000 0x0 0x00010000>; > }; > }; > > host1x@50000000 { > dc@54200000 { > memory-region = <&fb>, <&adsp>; > }; > }; > > This is abbreviated a little to focus on the essentials. Note also that > the ADSP reservation is not actually used on this device and the driver > for this doesn't exist yet, but I wanted to include this variant for > testing, because we'll want to use these bindings for the reservation > use-case as well at some point. > > I've also been able to make use of this binding and the IOMMU code in > conjunction with the simple-framebuffer driver to hand over a display > configuration set up by UEFI to the Linux kernel. > > Janne has confirmed[0] this to be suitable for indirect mappings as > well, though these patches don't implement that feature yet. Potential > extensions to this have been discussed but are not yet included at this > time to not further complicate things. > > Thierry > > [0]: https://lore.kernel.org/all/20220909144504.GA4024@jannau.net/ > > Thierry Reding (4): > of: Introduce of_translate_dma_region() > dt-bindings: reserved-memory: Document iommu-addresses > iommu: Implement of_iommu_get_resv_regions() > iommu: dma: Use of_iommu_get_resv_regions() > > .../reserved-memory/reserved-memory.yaml | 89 +++++++++++++++++- > drivers/iommu/dma-iommu.c | 3 + > drivers/iommu/of_iommu.c | 94 +++++++++++++++++++ > drivers/of/address.c | 41 ++++++++ > include/linux/of_address.h | 2 + > include/linux/of_iommu.h | 8 ++ > 6 files changed, 233 insertions(+), 4 deletions(-) Joerg, Rob, Is there anything left to do on the series? It'd be great to get some feedback from Robin on patch 3 since he had some concerns about how the reservation type was getting determined. All those should now be addressed and I think overall this is ready to go. Rob, you've given a Reviewed-by on all the DT-related parts, does that mean you're okay with this going through Joerg's tree? Joerg, other than a Reviewed-by from Robin on patch 3, anything else you'd like to see before you pick this up? Thierry
On Fri, Dec 2, 2022 at 8:41 AM Thierry Reding <thierry.reding@gmail.com> wrote: > > On Thu, Nov 17, 2022 at 07:54:20PM +0100, Thierry Reding wrote: > > From: Thierry Reding <treding@nvidia.com> > > > > Hi, > > > > This version is a minor update to the previous v11, which can be found > > here: > > > > https://lore.kernel.org/all/20221111161806.630527-1-thierry.reding@gmail.com/ > > > > The only change here is that the #dma-{address,size}-cells is dropped. > > It turns out to be much simpler to just update #{address,size}-cells to > > what they should be rather than add extra complexity for the DMA work- > > around. There's a minor update to the DT binding so that it can now > > properly validate cases where we have both reg and iommu-addresses > > properties. > > > > An example is included in the DT bindings, but here is an extract of > > what I've used to test this: > > > > reserved-memory { > > #address-cells = <2>; > > #size-cells = <2>; > > ranges; > > > > /* > > * Creates an identity mapping for the framebuffer that > > * the firmware has setup to scan out a bootsplash from. > > */ > > fb: framebuffer@92cb2000 { > > reg = <0x0 0x92cb2000 0x0 0x00800000>; > > iommu-addresses = <&dc0 0x0 0x92cb2000 0x0 0x00800000>; > > }; > > > > /* > > * Creates a reservation in the IOVA space to prevent > > * any buffers from being mapped to that region. Note > > * that on Tegra the range is actually quite different > > * from this, but it would conflict with the display > > * driver that I tested this against, so this is just > > * a dummy region for testing. > > */ > > adsp: reservation-adsp { > > iommu-addresses = <&dc0 0x0 0x90000000 0x0 0x00010000>; > > }; > > }; > > > > host1x@50000000 { > > dc@54200000 { > > memory-region = <&fb>, <&adsp>; > > }; > > }; > > > > This is abbreviated a little to focus on the essentials. Note also that > > the ADSP reservation is not actually used on this device and the driver > > for this doesn't exist yet, but I wanted to include this variant for > > testing, because we'll want to use these bindings for the reservation > > use-case as well at some point. > > > > I've also been able to make use of this binding and the IOMMU code in > > conjunction with the simple-framebuffer driver to hand over a display > > configuration set up by UEFI to the Linux kernel. > > > > Janne has confirmed[0] this to be suitable for indirect mappings as > > well, though these patches don't implement that feature yet. Potential > > extensions to this have been discussed but are not yet included at this > > time to not further complicate things. > > > > Thierry > > > > [0]: https://lore.kernel.org/all/20220909144504.GA4024@jannau.net/ > > > > Thierry Reding (4): > > of: Introduce of_translate_dma_region() > > dt-bindings: reserved-memory: Document iommu-addresses > > iommu: Implement of_iommu_get_resv_regions() > > iommu: dma: Use of_iommu_get_resv_regions() > > > > .../reserved-memory/reserved-memory.yaml | 89 +++++++++++++++++- > > drivers/iommu/dma-iommu.c | 3 + > > drivers/iommu/of_iommu.c | 94 +++++++++++++++++++ > > drivers/of/address.c | 41 ++++++++ > > include/linux/of_address.h | 2 + > > include/linux/of_iommu.h | 8 ++ > > 6 files changed, 233 insertions(+), 4 deletions(-) > > Joerg, Rob, > > Is there anything left to do on the series? It'd be great to get some > feedback from Robin on patch 3 since he had some concerns about how the > reservation type was getting determined. All those should now be > addressed and I think overall this is ready to go. > > Rob, you've given a Reviewed-by on all the DT-related parts, does that > mean you're okay with this going through Joerg's tree? Yes. Okay with and what I'm expecting. Rob
On Thu, Nov 17, 2022 at 07:54:20PM +0100, Thierry Reding wrote: > From: Thierry Reding <treding@nvidia.com> > > Hi, > > This version is a minor update to the previous v11, which can be found > here: > > https://lore.kernel.org/all/20221111161806.630527-1-thierry.reding@gmail.com/ > > The only change here is that the #dma-{address,size}-cells is dropped. > It turns out to be much simpler to just update #{address,size}-cells to > what they should be rather than add extra complexity for the DMA work- > around. There's a minor update to the DT binding so that it can now > properly validate cases where we have both reg and iommu-addresses > properties. > > An example is included in the DT bindings, but here is an extract of > what I've used to test this: > > reserved-memory { > #address-cells = <2>; > #size-cells = <2>; > ranges; > > /* > * Creates an identity mapping for the framebuffer that > * the firmware has setup to scan out a bootsplash from. > */ > fb: framebuffer@92cb2000 { > reg = <0x0 0x92cb2000 0x0 0x00800000>; > iommu-addresses = <&dc0 0x0 0x92cb2000 0x0 0x00800000>; > }; > > /* > * Creates a reservation in the IOVA space to prevent > * any buffers from being mapped to that region. Note > * that on Tegra the range is actually quite different > * from this, but it would conflict with the display > * driver that I tested this against, so this is just > * a dummy region for testing. > */ > adsp: reservation-adsp { > iommu-addresses = <&dc0 0x0 0x90000000 0x0 0x00010000>; > }; > }; > > host1x@50000000 { > dc@54200000 { > memory-region = <&fb>, <&adsp>; > }; > }; > > This is abbreviated a little to focus on the essentials. Note also that > the ADSP reservation is not actually used on this device and the driver > for this doesn't exist yet, but I wanted to include this variant for > testing, because we'll want to use these bindings for the reservation > use-case as well at some point. > > I've also been able to make use of this binding and the IOMMU code in > conjunction with the simple-framebuffer driver to hand over a display > configuration set up by UEFI to the Linux kernel. > > Janne has confirmed[0] this to be suitable for indirect mappings as > well, though these patches don't implement that feature yet. Potential > extensions to this have been discussed but are not yet included at this > time to not further complicate things. > > Thierry > > [0]: https://lore.kernel.org/all/20220909144504.GA4024@jannau.net/ > > Thierry Reding (4): > of: Introduce of_translate_dma_region() > dt-bindings: reserved-memory: Document iommu-addresses > iommu: Implement of_iommu_get_resv_regions() > iommu: dma: Use of_iommu_get_resv_regions() Hi Joerg, I think this has all the Acked-by's and Reviewed-by's that it needs. Can you pick this up? I'd really like to see this go into v6.3. If there's anything that you'd like to see addressed, please let me know. Thanks, Thierry
On Fri, Jan 20, 2023 at 06:36:15PM +0100, Thierry Reding wrote: > On Thu, Nov 17, 2022 at 07:54:20PM +0100, Thierry Reding wrote: > > From: Thierry Reding <treding@nvidia.com> > > > > Hi, > > > > This version is a minor update to the previous v11, which can be found > > here: > > > > https://lore.kernel.org/all/20221111161806.630527-1-thierry.reding@gmail.com/ > > > > The only change here is that the #dma-{address,size}-cells is dropped. > > It turns out to be much simpler to just update #{address,size}-cells to > > what they should be rather than add extra complexity for the DMA work- > > around. There's a minor update to the DT binding so that it can now > > properly validate cases where we have both reg and iommu-addresses > > properties. > > > > An example is included in the DT bindings, but here is an extract of > > what I've used to test this: > > > > reserved-memory { > > #address-cells = <2>; > > #size-cells = <2>; > > ranges; > > > > /* > > * Creates an identity mapping for the framebuffer that > > * the firmware has setup to scan out a bootsplash from. > > */ > > fb: framebuffer@92cb2000 { > > reg = <0x0 0x92cb2000 0x0 0x00800000>; > > iommu-addresses = <&dc0 0x0 0x92cb2000 0x0 0x00800000>; > > }; > > > > /* > > * Creates a reservation in the IOVA space to prevent > > * any buffers from being mapped to that region. Note > > * that on Tegra the range is actually quite different > > * from this, but it would conflict with the display > > * driver that I tested this against, so this is just > > * a dummy region for testing. > > */ > > adsp: reservation-adsp { > > iommu-addresses = <&dc0 0x0 0x90000000 0x0 0x00010000>; > > }; > > }; > > > > host1x@50000000 { > > dc@54200000 { > > memory-region = <&fb>, <&adsp>; > > }; > > }; > > > > This is abbreviated a little to focus on the essentials. Note also that > > the ADSP reservation is not actually used on this device and the driver > > for this doesn't exist yet, but I wanted to include this variant for > > testing, because we'll want to use these bindings for the reservation > > use-case as well at some point. > > > > I've also been able to make use of this binding and the IOMMU code in > > conjunction with the simple-framebuffer driver to hand over a display > > configuration set up by UEFI to the Linux kernel. > > > > Janne has confirmed[0] this to be suitable for indirect mappings as > > well, though these patches don't implement that feature yet. Potential > > extensions to this have been discussed but are not yet included at this > > time to not further complicate things. > > > > Thierry > > > > [0]: https://lore.kernel.org/all/20220909144504.GA4024@jannau.net/ > > > > Thierry Reding (4): > > of: Introduce of_translate_dma_region() > > dt-bindings: reserved-memory: Document iommu-addresses > > iommu: Implement of_iommu_get_resv_regions() > > iommu: dma: Use of_iommu_get_resv_regions() > > Hi Joerg, > > I think this has all the Acked-by's and Reviewed-by's that it needs. Can > you pick this up? I'd really like to see this go into v6.3. > > If there's anything that you'd like to see addressed, please let me > know. Joerg, I just noticed that this no longer applies on v6.2-rc4, so I just sent out a v13 which is really only a rebase with the tags from v12 applied, no further changes. Again, let me know if there's anything else you think this needs before it can be picked up. Thanks, Thierry
From: Thierry Reding <treding@nvidia.com> Hi, This version is a minor update to the previous v11, which can be found here: https://lore.kernel.org/all/20221111161806.630527-1-thierry.reding@gmail.com/ The only change here is that the #dma-{address,size}-cells is dropped. It turns out to be much simpler to just update #{address,size}-cells to what they should be rather than add extra complexity for the DMA work- around. There's a minor update to the DT binding so that it can now properly validate cases where we have both reg and iommu-addresses properties. An example is included in the DT bindings, but here is an extract of what I've used to test this: reserved-memory { #address-cells = <2>; #size-cells = <2>; ranges; /* * Creates an identity mapping for the framebuffer that * the firmware has setup to scan out a bootsplash from. */ fb: framebuffer@92cb2000 { reg = <0x0 0x92cb2000 0x0 0x00800000>; iommu-addresses = <&dc0 0x0 0x92cb2000 0x0 0x00800000>; }; /* * Creates a reservation in the IOVA space to prevent * any buffers from being mapped to that region. Note * that on Tegra the range is actually quite different * from this, but it would conflict with the display * driver that I tested this against, so this is just * a dummy region for testing. */ adsp: reservation-adsp { iommu-addresses = <&dc0 0x0 0x90000000 0x0 0x00010000>; }; }; host1x@50000000 { dc@54200000 { memory-region = <&fb>, <&adsp>; }; }; This is abbreviated a little to focus on the essentials. Note also that the ADSP reservation is not actually used on this device and the driver for this doesn't exist yet, but I wanted to include this variant for testing, because we'll want to use these bindings for the reservation use-case as well at some point. I've also been able to make use of this binding and the IOMMU code in conjunction with the simple-framebuffer driver to hand over a display configuration set up by UEFI to the Linux kernel. Janne has confirmed[0] this to be suitable for indirect mappings as well, though these patches don't implement that feature yet. Potential extensions to this have been discussed but are not yet included at this time to not further complicate things. Thierry [0]: https://lore.kernel.org/all/20220909144504.GA4024@jannau.net/ Thierry Reding (4): of: Introduce of_translate_dma_region() dt-bindings: reserved-memory: Document iommu-addresses iommu: Implement of_iommu_get_resv_regions() iommu: dma: Use of_iommu_get_resv_regions() .../reserved-memory/reserved-memory.yaml | 89 +++++++++++++++++- drivers/iommu/dma-iommu.c | 3 + drivers/iommu/of_iommu.c | 94 +++++++++++++++++++ drivers/of/address.c | 41 ++++++++ include/linux/of_address.h | 2 + include/linux/of_iommu.h | 8 ++ 6 files changed, 233 insertions(+), 4 deletions(-)