Linaro-mm-sig

linaro-mm-sig@lists.linaro.org

15 participants
3030 discussions

Re: [Linaro-mm-sig] New "xf86-video-armsoc" DDX driver

by Dave Airlie

> > For the last few months we (ARM MPD... "The Mali guys") have been working on > getting X.Org up and running with Mali T6xx (ARM's next-generation GPU IP). > The approach is very similar (well identical I think) to how things work on > OMAP: We use a DRM driver to manage the display controller via KMS. The KMS > driver also allocates both scan-out and pixmap/back buffers via the > DRM_IOCTL_MODE_CREATE_DUMB ioctl which is internally implemented with GEM. > When returning buffers to DRI clients, the x-server uses flink to get a > global handle to a buffer which it passes back to the DRI client (in our > case the Mali-T600 X11 EGL winsys). The client then uses the new PRIME > ioctls to export the GEM buffer it received from the x-server to a dma_buf > fd. This fd is then passed into the T6xx kernel driver via our own job > dispatch user/kernel API (we're not using DRM for driving the GPU, only the > display controller). So using dumb in this was is probably a bit of an abuse, since dumb is defined to provide buffers not to be used for acceleration hw. Since when we allocate dumb buffers, we can't know what special hw layouts are required (tiling etc) for optimal performance for accel. The logic to work that out is rarely generic. > > http://git.linaro.org/gitweb?p=arm/xorg/driver/xf86-video-armsoc.git;a=summa > ry > > Note: When we originally spoke to Rob Clark about this, he suggested we take > the already-generic xf86-video-modesetting and just add the dri2 code to it. > This is indeed how we started out, however as we progressed it became clear > that the majority of the code we wanted was in the omap driver and were > having to work fairly hard to keep some of the original modesetting code. > This is why we've now changed tactic and just forked the OMAP driver, > something Rob is more than happy for us to do. It does seem like porting to -modesetting, and maybe cleaning up modesetting if its needs it. The modesetting driver is pretty much just a make it work port of the radeon/nouveau/intel code "shared" code. > One thing the DDX driver isn't doing yet is making use of 2D hw blocks. In > the short-term, we will simply create a branch off of the "generic" master > for each SoC and add 2D hardware support there. We do however want a more > permanent solution which doesn't need a separate branch per SoC. Some of the > suggested solutions are: > > * Add a new generic DRM ioctl API for larger 2D operations (I would imagine > small blits/blends would be done in SW). Not going to happen, again the hw isn't generic in this area, some hw requires 3D engines to do 2D ops etc. The limitations on some hw with overlaps etc, and finally it breaks the rule about generic ioctls for acceleration operations. > * Use SW rendering for everything other than solid blits and use v4l2's > blitting API for those (importing/exporting buffers to be blitted using > dma_buf). The theory here is that most UIs are rendered with GLES and so you > only need 2D hardware for blits. I think we'll prototype this approach on > Exynos. Seems a bit over the top, > * Define a new x-server sub-module interface to allow a seperate .so 2D > driver to be loaded (this is the approach the current OMAP DDX uses). This seems the sanest. I haven't time this week to review the code, but I'll try and take a look when time permits. Dave.

13 years, 3 months

[PATCH] dma-buf: minor documentation fixes.

by Sumit Semwal

Some minor inline documentation fixes for gaps resulting from new patches. Signed-off-by: Sumit Semwal <sumit.semwal(a)ti.com> --- drivers/base/dma-buf.c | 9 +++++---- include/linux/dma-buf.h | 3 +++ 2 files changed, 8 insertions(+), 4 deletions(-) diff --git a/drivers/base/dma-buf.c b/drivers/base/dma-buf.c index c3c88b0..24e88fe 100644 --- a/drivers/base/dma-buf.c +++ b/drivers/base/dma-buf.c @@ -429,7 +429,7 @@ EXPORT_SYMBOL_GPL(dma_buf_kunmap); /** * dma_buf_mmap - Setup up a userspace mmap with the given vma - * @dma_buf: [in] buffer that should back the vma + * @dmabuf: [in] buffer that should back the vma * @vma: [in] vma for the mmap * @pgoff: [in] offset in pages where this mmap should start within the * dma-buf buffer. @@ -470,8 +470,9 @@ int dma_buf_mmap(struct dma_buf *dmabuf, struct vm_area_struct *vma, EXPORT_SYMBOL_GPL(dma_buf_mmap); /** - * dma_buf_vmap - Create virtual mapping for the buffer object into kernel address space. Same restrictions as for vmap and friends apply. - * @dma_buf: [in] buffer to vmap + * dma_buf_vmap - Create virtual mapping for the buffer object into kernel + * address space. Same restrictions as for vmap and friends apply. + * @dmabuf: [in] buffer to vmap * * This call may fail due to lack of virtual mapping address space. * These calls are optional in drivers. The intended use for them @@ -491,7 +492,7 @@ EXPORT_SYMBOL_GPL(dma_buf_vmap); /** * dma_buf_vunmap - Unmap a vmap obtained by dma_buf_vmap. - * @dma_buf: [in] buffer to vmap + * @dmabuf: [in] buffer to vunmap */ void dma_buf_vunmap(struct dma_buf *dmabuf, void *vaddr) { diff --git a/include/linux/dma-buf.h b/include/linux/dma-buf.h index a02b1ff..eb48f38 100644 --- a/include/linux/dma-buf.h +++ b/include/linux/dma-buf.h @@ -65,6 +65,9 @@ struct dma_buf_attachment; * mapping needs to be coherent - if the exporter doesn't directly * support this, it needs to fake coherency by shooting down any ptes * when transitioning away from the cpu domain. + * @vmap: [optional] creates a virtual mapping for the buffer into kernel + * address space. Same restrictions as for vmap and friends apply. + * @vunmap: [optional] unmaps a vmap from the buffer */ struct dma_buf_ops { int (*attach)(struct dma_buf *, struct device *, -- 1.7.9.5

13 years, 3 months

[PATCH] dma-buf: add get_dma_buf()

by Rob Clark

From: Rob Clark <rob(a)ti.com> Works in a similar way to get_file(), and is needed in cases such as when the exporter needs to also keep a reference to the dmabuf (that is later released with a dma_buf_put()), and possibly other similar cases. Signed-off-by: Rob Clark <rob(a)ti.com> --- include/linux/dma-buf.h | 14 ++++++++++++++ 1 files changed, 14 insertions(+), 0 deletions(-) diff --git a/include/linux/dma-buf.h b/include/linux/dma-buf.h index cbdff81..25eb287 100644 --- a/include/linux/dma-buf.h +++ b/include/linux/dma-buf.h @@ -132,6 +132,20 @@ struct dma_buf_attachment { void *priv; }; +/** + * get_dma_buf - convenience wrapper for get_file. + * @dmabuf: [in] pointer to dma_buf + * + * Increments the reference count on the dma-buf, needed in case of drivers + * that either need to create additional references to the dmabuf on the + * kernel side. For example, an exporter that needs to keep a dmabuf ptr + * so that subsequent exports don't create a new dmabuf. + */ +static inline void get_dma_buf(struct dma_buf *dmabuf) +{ + get_file(dmabuf->file); +} + #ifdef CONFIG_DMA_SHARED_BUFFER struct dma_buf_attachment *dma_buf_attach(struct dma_buf *dmabuf, struct device *dev); -- 1.7.5.4

13 years, 3 months

[PATCH] dma-buf: add initial vmap documentation

by Dave Airlie

From: Dave Airlie <airlied(a)redhat.com> Signed-off-by: Dave Airlie <airlied(a)redhat.com> --- Documentation/dma-buf-sharing.txt | 11 +++++++++++ 1 files changed, 11 insertions(+), 0 deletions(-) diff --git a/Documentation/dma-buf-sharing.txt b/Documentation/dma-buf-sharing.txt index 3bbd5c5..98e9fa0 100644 --- a/Documentation/dma-buf-sharing.txt +++ b/Documentation/dma-buf-sharing.txt @@ -300,6 +300,17 @@ Access to a dma_buf from the kernel context involves three steps: Note that these calls need to always succeed. The exporter needs to complete any preparations that might fail in begin_cpu_access. + For some circumstances the overhead of kmap can be too high, a vmap interface + is introduced. This interface shouldn't be used very carefully, as vmalloc + space is a limited resources on many architectures. + + Interfaces: + void *dma_buf_vmap(struct dma_buf *dmabuf) + void dma_buf_vunmap(struct dma_buf *dmabuf, void *vaddr) + + This call can fail if there is no vmap support in the exporter, or if it + runs out of vmalloc space. Fallback to kmap should be implemented. + 3. Finish access When the importer is done accessing the range specified in begin_cpu_access, -- 1.7.6

13 years, 3 months

[RFC 00/13] Support for dmabuf exporting for videobuf2

by Tomasz Stanislawski

Hello everyone, The patches adds support for DMABUF [1] exporting to V4L2 stack. It was updated from [7] after Laurent Pinchart's review. The previous patchset was split into two parts. The support for DMABUF importing was posted in [2]. The exporter part is dependant on DMA mapping redesign [3] which is not merged into the mainline. Therefore it is posted as a separate patchset. This patchset is rebased on 3.4-rc1 plus the following patchsets: - support for DMABUF importing in V4L2 [2] - DMA mapping redesign [3] - support for dma_get_pages extension [4] - support for vmap extension to dmabuf framework by Dave Airlie [5][6] [1] https://lkml.org/lkml/2011/12/26/29 [2] http://thread.gmane.org/gmane.linux.drivers.video-input-infrastructure/46586 [3] http://thread.gmane.org/gmane.linux.kernel.cross-arch/12819 [4] http://thread.gmane.org/gmane.linux.drivers.video-input-infrastructure/4379… [5] http://thread.gmane.org/gmane.linux.drivers.video-input-infrastructure/46713 [6] http://cgit.freedesktop.org/~airlied/linux/commit/?h=drm-dmabuf2&id=c481a54… [7] http://thread.gmane.org/gmane.comp.video.dri.devel/66213 Marek Szyprowski (1): v4l: vb2-dma-contig: let mmap method to use dma_mmap_coherent call Tomasz Stanislawski (12): v4l: add buffer exporting via dmabuf v4l: vb2: add buffer exporting via dmabuf v4l: vb2-dma-contig: add setup of sglist for MMAP buffers v4l: vb2-dma-contig: add support for DMABUF exporting v4l: vb2-dma-contig: add vmap/kmap for dmabuf exporting v4l: vb2-dma-contig: change map/unmap behaviour for importers v4l: vb2-dma-contig: change map/unmap behaviour for exporters v4l: s5p-tv: mixer: support for dmabuf importing v4l: s5p-tv: mixer: support for dmabuf exporting v4l: fimc: support for dmabuf importing v4l: fimc: support for dmabuf exporting v4l: vivi: support for dmabuf exporting drivers/media/video/Kconfig | 1 + drivers/media/video/s5p-fimc/fimc-capture.c | 11 ++- drivers/media/video/s5p-tv/Kconfig | 1 + drivers/media/video/s5p-tv/mixer_video.c | 12 ++- drivers/media/video/v4l2-compat-ioctl32.c | 1 + drivers/media/video/v4l2-ioctl.c | 7 + drivers/media/video/videobuf2-core.c | 66 ++++++++ drivers/media/video/videobuf2-dma-contig.c | 224 +++++++++++++++++++++++++- drivers/media/video/vivi.c | 9 + include/linux/videodev2.h | 23 +++ include/media/v4l2-ioctl.h | 2 + include/media/videobuf2-core.h | 2 + 12 files changed, 348 insertions(+), 11 deletions(-) -- 1.7.5.4

13 years, 3 months

New "xf86-video-armsoc" DDX driver

by Tom Cooksey

Hi All, For the last few months we (ARM MPD... "The Mali guys") have been working on getting X.Org up and running with Mali T6xx (ARM's next-generation GPU IP). The approach is very similar (well identical I think) to how things work on OMAP: We use a DRM driver to manage the display controller via KMS. The KMS driver also allocates both scan-out and pixmap/back buffers via the DRM_IOCTL_MODE_CREATE_DUMB ioctl which is internally implemented with GEM. When returning buffers to DRI clients, the x-server uses flink to get a global handle to a buffer which it passes back to the DRI client (in our case the Mali-T600 X11 EGL winsys). The client then uses the new PRIME ioctls to export the GEM buffer it received from the x-server to a dma_buf fd. This fd is then passed into the T6xx kernel driver via our own job dispatch user/kernel API (we're not using DRM for driving the GPU, only the display controller). Note: ARM doesn't generally provide the display controller IP block, so this is really for our customers/Linaro to develop, though we do have something hacked up for ARM's own PL111 display controller on our Versatile Express development platform which we'll be open sourcing/up-streaming asap via Linaro. We believe most ARM SoCs are likely to work the same way, at least those with 3rd-party GPU IP blocks/drivers (so everyone except Qualcomm & nVidia). As mentioned, this is certainly how the OMAP integration works. As such, we've taken the OMAP DDX driver Rob Clark wrote and hacked on it to make it work for Mali. The patch is actually relatively small, which is not really too surprising as all the driver is doing is allocating buffers and managing a display controller via a device-agnostic interface (KMS). All the device-specific code is kept in the DRM driver and the client GLES/EGL library. Given that the DDX driver doesn't contain any device-specific code, we're going to take the OMAP DDX as a baseline and try and make it more generic. Our immediate goals are to make it work on our own Versatile Express development platform and on Samsung's Exynos 5250 SoC, however our hope is to have a single DDX driver which can cover OMAP, Exynos, ST-E's Nova/Thor platforms and probably others too. It's even been suggested it could work with Mesa's sw backend(?). Anyway, the DDX is very much a work-in-progress and is still heavily branded OMAP, even though it's working (almost) perfectly on VExpress & Exynos too (re-branding isn't too high-up our priority list at the moment). We are actively developing this driver and will be doing so in a public git repository hosted by Linaro. We will not be maintaining any private repository behind ARM's firewall or anything like that - you'll see what we see. The first patches have now been pushed, so if anyone's interested in seeing what we have so far or wants to track development, the tree is here: http://git.linaro.org/gitweb?p=arm/xorg/driver/xf86-video-armsoc.git;a=summa ry Note: When we originally spoke to Rob Clark about this, he suggested we take the already-generic xf86-video-modesetting and just add the dri2 code to it. This is indeed how we started out, however as we progressed it became clear that the majority of the code we wanted was in the omap driver and were having to work fairly hard to keep some of the original modesetting code. This is why we've now changed tactic and just forked the OMAP driver, something Rob is more than happy for us to do. One thing the DDX driver isn't doing yet is making use of 2D hw blocks. In the short-term, we will simply create a branch off of the "generic" master for each SoC and add 2D hardware support there. We do however want a more permanent solution which doesn't need a separate branch per SoC. Some of the suggested solutions are: * Add a new generic DRM ioctl API for larger 2D operations (I would imagine small blits/blends would be done in SW). * Use SW rendering for everything other than solid blits and use v4l2's blitting API for those (importing/exporting buffers to be blitted using dma_buf). The theory here is that most UIs are rendered with GLES and so you only need 2D hardware for blits. I think we'll prototype this approach on Exynos. * Define a new x-server sub-module interface to allow a seperate .so 2D driver to be loaded (this is the approach the current OMAP DDX uses). We are hoping someone might have some advice & suggestions on how to proceed with regards to 2D. We're also very interested in any feedback, both on the DDX driver specifically and on the approach we're taking in general. Cheers, Tom

13 years, 3 months

[PATCH] dma-buf: add vmap interface (v3)

by Dave Airlie

From: Dave Airlie <airlied(a)redhat.com> The main requirement I have for this interface is for scanning out using the USB gpu devices. Since these devices have to read the framebuffer on updates and linearly compress it, using kmaps is a major overhead for every update. v2: fix warn issues pointed out by Sylwester Nawrocki. v3: fix compile !CONFIG_DMA_SHARED_BUFFER and add _GPL for now Signed-off-by: Dave Airlie <airlied(a)redhat.com> --- drivers/base/dma-buf.c | 34 ++++++++++++++++++++++++++++++++++ include/linux/dma-buf.h | 14 ++++++++++++++ 2 files changed, 48 insertions(+), 0 deletions(-) diff --git a/drivers/base/dma-buf.c b/drivers/base/dma-buf.c index 07cbbc6..0d8197e 100644 --- a/drivers/base/dma-buf.c +++ b/drivers/base/dma-buf.c @@ -406,3 +406,37 @@ void dma_buf_kunmap(struct dma_buf *dmabuf, unsigned long page_num, dmabuf->ops->kunmap(dmabuf, page_num, vaddr); } EXPORT_SYMBOL_GPL(dma_buf_kunmap); + +/** + * dma_buf_vmap - Create virtual mapping for the buffer object into kernel address space. The same restrictions as for vmap and friends apply. + * @dma_buf: [in] buffer to vmap + * + * This call may fail due to lack of virtual mapping address space. + * These calls are optional in drivers. The intended use for them + * is for mapping objects linear in kernel space for high use objects. + * Please attempt to use kmap/kunmap before thinking about these interfaces. + */ +void *dma_buf_vmap(struct dma_buf *dmabuf) +{ + if (WARN_ON(!dmabuf)) + return NULL; + + if (dmabuf->ops->vmap) + return dmabuf->ops->vmap(dmabuf); + return NULL; +} +EXPORT_SYMBOL_GPL(dma_buf_vmap); + +/** + * dma_buf_vunmap - Unmap a vmap obtained by dma_buf_vmap. + * @dma_buf: [in] buffer to vmap + */ +void dma_buf_vunmap(struct dma_buf *dmabuf, void *vaddr) +{ + if (WARN_ON(!dmabuf)) + return; + + if (dmabuf->ops->vunmap) + dmabuf->ops->vunmap(dmabuf, vaddr); +} +EXPORT_SYMBOL_GPL(dma_buf_vunmap); diff --git a/include/linux/dma-buf.h b/include/linux/dma-buf.h index 3efbfc2..d8c2865 100644 --- a/include/linux/dma-buf.h +++ b/include/linux/dma-buf.h @@ -92,6 +92,9 @@ struct dma_buf_ops { void (*kunmap_atomic)(struct dma_buf *, unsigned long, void *); void *(*kmap)(struct dma_buf *, unsigned long); void (*kunmap)(struct dma_buf *, unsigned long, void *); + + void *(*vmap)(struct dma_buf *); + void (*vunmap)(struct dma_buf *, void *vaddr); }; /** @@ -167,6 +170,9 @@ void *dma_buf_kmap_atomic(struct dma_buf *, unsigned long); void dma_buf_kunmap_atomic(struct dma_buf *, unsigned long, void *); void *dma_buf_kmap(struct dma_buf *, unsigned long); void dma_buf_kunmap(struct dma_buf *, unsigned long, void *); + +void *dma_buf_vmap(struct dma_buf *); +void dma_buf_vunmap(struct dma_buf *, void *vaddr); #else static inline struct dma_buf_attachment *dma_buf_attach(struct dma_buf *dmabuf, @@ -248,6 +254,14 @@ static inline void dma_buf_kunmap(struct dma_buf *dmabuf, unsigned long pnum, void *vaddr) { } + +static inline void *dma_buf_vmap(struct dma_buf *dmabuf) +{ +} + +static inline void dma_buf_vunmap(struct dma_buf *dmabuf, void *vaddr) +{ +} #endif /* CONFIG_DMA_SHARED_BUFFER */ #endif /* __DMA_BUF_H__ */ -- 1.7.6

13 years, 3 months

[RFC 0/2] dma-mapping: Introduce new IOVA API with address specified

by Hiroshi DOYU

Hello, The following patchset is our enhancement for the upstream DMA mapping API(v9), where new IOVA API is introduced with the version of IOVA address specified. The current upstream DMA mapping API cannot specify any specific IOVA address at allocation. We need to specify IOVA address. This is necessary because some HWAs requre some specific address, for example, AVP vector and also some data buffer alignement can improve better performance from H/W constraints POV. Hiroshi DOYU (2): dma-mapping: Export arm_iommu_{alloc,free}_iova() functions dma-mapping: Enable IOVA mapping with specific address arch/arm/include/asm/dma-iommu.h | 31 ++++++ arch/arm/include/asm/dma-mapping.h | 1 + arch/arm/mm/dma-mapping.c | 181 +++++++++++++++++++++++++++++------- 3 files changed, 180 insertions(+), 33 deletions(-) -- 1.7.5.4

13 years, 4 months

[PATCH] dma-buf: mmap support

by Daniel Vetter

Compared to Rob Clark's RFC I've ditched the prepare/finish hooks and corresponding ioctls on the dma_buf file. The major reason for that is that many people seem to be under the impression that this is also for synchronization with outstanding asynchronous processsing. I'm pretty massively opposed to this because: - It boils down reinventing a new rather general-purpose userspace synchronization interface. If we look at things like futexes, this is hard to get right. - Furthermore a lot of kernel code has to interact with this synchronization primitive. This smells a look like the dri1 hw_lock, a horror show I prefer not to reinvent. - Even more fun is that multiple different subsystems would interact here, so we have plenty of opportunities to create funny deadlock scenarios. I think synchronization is a wholesale different problem from data sharing and should be tackled as an orthogonal problem. Now we could demand that prepare/finish may only ensure cache coherency (as Rob intended), but that runs up into the next problem: We not only need mmap support to facilitate sw-only processing nodes in a pipeline (without jumping through hoops by importing the dma_buf into some sw-access only importer), which allows for a nicer ION->dma-buf upgrade path for existing Android userspace. We also need mmap support for existing importing subsystems to support existing userspace libraries. And a loot of these subsystems are expected to export coherent userspace mappings. So prepare/finish can only ever be optional and the exporter /needs/ to support coherent mappings. Given that mmap access is always somewhat fallback-y in nature I've decided to drop this optimization, instead of just making it optional. If we demonstrate a clear need for this, supported by benchmark results, we can always add it in again later as an optional extension. Other differences compared to Rob's RFC is the above mentioned support for mapping a dma-buf through facilities provided by the importer. Which results in mmap support no longer being optional. Note that this dma-buf mmap patch does _not_ support every possible insanity an existing subsystem could pull of with mmap: Because it does not allow to intercept pagefaults and shoot down ptes importing subsystems can't add some magic of their own at these points (e.g. to automatically synchronize with outstanding rendering or set up some special resources). I've done a cursory read through a few mmap implementions of various subsytems and I'm hopeful that we can avoid this (and the complexity it'd bring with it). Additonally I've extended the documentation a bit to explain the hows and whys of this mmap extension. In case we ever want to add support for explicitly cache maneged userspace mmap with a prepare/finish ioctl pair, we could specify that userspace needs to mmap a different part of the dma_buf, e.g. the range starting at dma_buf->size up to dma_buf->size*2. This works because the size of a dma_buf is invariant over it's lifetime. The exporter would obviously need to fall back to coherent mappings for both ranges if a legacy clients maps the coherent range and the architecture cannot suppor conflicting caching policies. Also, this would obviously be optional and userspace needs to be able to fall back to coherent mappings. v2: - Spelling fixes from Rob Clark. - Compile fix for !DMA_BUF from Rob Clark. - Extend commit message to explain how explicitly cache managed mmap support could be added later. - Extend the documentation with implementations notes for exporters that need to manually fake coherency. v3: - dma_buf pointer initialization goof-up noticed by Rebecca Schultz Zavin. Cc: Rob Clark <rob.clark(a)linaro.org> Cc: Rebecca Schultz Zavin <rebecca(a)android.com> Signed-Off-by: Daniel Vetter <daniel.vetter(a)ffwll.ch> --- Documentation/dma-buf-sharing.txt | 98 ++++++++++++++++++++++++++++++++++--- drivers/base/dma-buf.c | 64 +++++++++++++++++++++++- include/linux/dma-buf.h | 16 ++++++ 3 files changed, 170 insertions(+), 8 deletions(-) diff --git a/Documentation/dma-buf-sharing.txt b/Documentation/dma-buf-sharing.txt index 3bbd5c5..5ff4d2b 100644 --- a/Documentation/dma-buf-sharing.txt +++ b/Documentation/dma-buf-sharing.txt @@ -29,13 +29,6 @@ The buffer-user in memory, mapped into its own address space, so it can access the same area of memory. -*IMPORTANT*: [see https://lkml.org/lkml/2011/12/20/211 for more details] -For this first version, A buffer shared using the dma_buf sharing API: -- *may* be exported to user space using "mmap" *ONLY* by exporter, outside of - this framework. -- with this new iteration of the dma-buf api cpu access from the kernel has been - enable, see below for the details. - dma-buf operations for device dma only -------------------------------------- @@ -313,6 +306,83 @@ Access to a dma_buf from the kernel context involves three steps: enum dma_data_direction dir); +Direct Userspace Access/mmap Support +------------------------------------ + +Being able to mmap an export dma-buf buffer object has 2 main use-cases: +- CPU fallback processing in a pipeline and +- supporting existing mmap interfaces in importers. + +1. CPU fallback processing in a pipeline + + In many processing pipelines it is sometimes required that the cpu can access + the data in a dma-buf (e.g. for thumbnail creation, snapshots, ...). To avoid + the need to handle this specially in userspace frameworks for buffer sharing + it's ideal if the dma_buf fd itself can be used to access the backing storage + from userspace using mmap. + + Furthermore Android's ION framework already supports this (and is otherwise + rather similar to dma-buf from a userspace consumer side with using fds as + handles, too). So it's beneficial to support this in a similar fashion on + dma-buf to have a good transition path for existing Android userspace. + + No special interfaces, userspace simply calls mmap on the dma-buf fd. + +2. Supporting existing mmap interfaces in exporters + + Similar to the motivation for kernel cpu access it is again important that + the userspace code of a given importing subsystem can use the same interfaces + with a imported dma-buf buffer object as with a native buffer object. This is + especially important for drm where the userspace part of contemporary OpenGL, + X, and other drivers is huge, and reworking them to use a different way to + mmap a buffer rather invasive. + + The assumption in the current dma-buf interfaces is that redirecting the + initial mmap is all that's needed. A survey of some of the existing + subsystems shows that no driver seems to do any nefarious thing like syncing + up with outstanding asynchronous processing on the device or allocating + special resources at fault time. So hopefully this is good enough, since + adding interfaces to intercept pagefaults and allow pte shootdowns would + increase the complexity quite a bit. + + Interface: + int dma_buf_mmap(struct dma_buf *, struct vm_area_struct *, + unsigned long); + + If the importing subsystem simply provides a special-purpose mmap call to set + up a mapping in userspace, calling do_mmap with dma_buf->file will equally + achieve that for a dma-buf object. + +3. Implementation notes for exporters + + Because dma-buf buffers have invariant size over their lifetime, the dma-buf + core checks whether a vma is too large and rejects such mappings. The + exporter hence does not need to duplicate this check. + + Because existing importing subsystems might presume coherent mappings for + userspace, the exporter needs to set up a coherent mapping. If that's not + possible, it needs to fake coherency by manually shooting down ptes when + leaving the cpu domain and flushing caches at fault time. Note that all the + dma_buf files share the same anon inode, hence the exporter needs to replace + the dma_buf file stored in vma->vm_file with it's own if pte shootdown is + requred. This is because the kernel uses the underlying inode's address_space + for vma tracking (and hence pte tracking at shootdown time with + unmap_mapping_range). + + If the above shootdown dance turns out to be too expensive in certain + scenarios, we can extend dma-buf with a more explicit cache tracking scheme + for userspace mappings. But the current assumption is that using mmap is + always a slower path, so some inefficiencies should be acceptable. + + Exporters that shoot down mappings (for any reasons) shall not do any + synchronization at fault time with outstanding device operations. + Synchronization is an orthogonal issue to sharing the backing storage of a + buffer and hence should not be handled by dma-buf itself. This is explictly + mentioned here because many people seem to want something like this, but if + different exporters handle this differently, buffer sharing can fail in + interesting ways depending upong the exporter (if userspace starts depending + upon this implicit synchronization). + Miscellaneous notes ------------------- @@ -336,6 +406,20 @@ Miscellaneous notes the exporting driver to create a dmabuf fd must provide a way to let userspace control setting of O_CLOEXEC flag passed in to dma_buf_fd(). +- If an exporter needs to manually flush caches and hence needs to fake + coherency for mmap support, it needs to be able to zap all the ptes pointing + at the backing storage. Now linux mm needs a struct address_space associated + with the struct file stored in vma->vm_file to do that with the function + unmap_mapping_range. But the dma_buf framework only backs every dma_buf fd + with the anon_file struct file, i.e. all dma_bufs share the same file. + + Hence exporters need to setup their own file (and address_space) association + by setting vma->vm_file and adjusting vma->vm_pgoff in the dma_buf mmap + callback. In the specific case of a gem driver the exporter could use the + shmem file already provided by gem (and set vm_pgoff = 0). Exporters can then + zap ptes by unmapping the corresponding range of the struct address_space + associated with their own file. + References: [1] struct dma_buf_ops in include/linux/dma-buf.h [2] All interfaces mentioned above defined in include/linux/dma-buf.h diff --git a/drivers/base/dma-buf.c b/drivers/base/dma-buf.c index 07cbbc6..7cfb405 100644 --- a/drivers/base/dma-buf.c +++ b/drivers/base/dma-buf.c @@ -44,8 +44,26 @@ static int dma_buf_release(struct inode *inode, struct file *file) return 0; } +static int dma_buf_mmap_internal(struct file *file, struct vm_area_struct *vma) +{ + struct dma_buf *dmabuf; + + if (!is_dma_buf_file(file)) + return -EINVAL; + + dmabuf = file->private_data; + + /* check for overflowing the buffer's size */ + if (vma->vm_pgoff + ((vma->vm_end - vma->vm_start) >> PAGE_SHIFT) > + dmabuf->size >> PAGE_SHIFT) + return -EINVAL; + + return dmabuf->ops->mmap(dmabuf, vma); +} + static const struct file_operations dma_buf_fops = { .release = dma_buf_release, + .mmap = dma_buf_mmap_internal, }; /* @@ -82,7 +100,8 @@ struct dma_buf *dma_buf_export(void *priv, const struct dma_buf_ops *ops, || !ops->unmap_dma_buf || !ops->release || !ops->kmap_atomic - || !ops->kmap)) { + || !ops->kmap + || !ops->mmap)) { return ERR_PTR(-EINVAL); } @@ -406,3 +425,46 @@ void dma_buf_kunmap(struct dma_buf *dmabuf, unsigned long page_num, dmabuf->ops->kunmap(dmabuf, page_num, vaddr); } EXPORT_SYMBOL_GPL(dma_buf_kunmap); + + +/** + * dma_buf_mmap - Setup up a userspace mmap with the given vma + * @dma_buf: [in] buffer that should back the vma + * @vma: [in] vma for the mmap + * @pgoff: [in] offset in pages where this mmap should start within the + * dma-buf buffer. + * + * This function adjusts the passed in vma so that it points at the file of the + * dma_buf operation. It alsog adjusts the starting pgoff and does bounds + * checking on the size of the vma. Then it calls the exporters mmap function to + * set up the mapping. + * + * Can return negative error values, returns 0 on success. + */ +int dma_buf_mmap(struct dma_buf *dmabuf, struct vm_area_struct *vma, + unsigned long pgoff) +{ + if (WARN_ON(!dmabuf || !vma)) + return -EINVAL; + + /* check for offset overflow */ + if (pgoff + ((vma->vm_end - vma->vm_start) >> PAGE_SHIFT) < pgoff) + return -EOVERFLOW; + + /* check for overflowing the buffer's size */ + if (pgoff + ((vma->vm_end - vma->vm_start) >> PAGE_SHIFT) > + dmabuf->size >> PAGE_SHIFT) + return -EINVAL; + + /* readjust the vma */ + if (vma->vm_file) + fput(vma->vm_file); + + vma->vm_file = dmabuf->file; + get_file(vma->vm_file); + + vma->vm_pgoff = pgoff; + + return dmabuf->ops->mmap(dmabuf, vma); +} +EXPORT_SYMBOL_GPL(dma_buf_mmap); diff --git a/include/linux/dma-buf.h b/include/linux/dma-buf.h index 3efbfc2..1f78d15 100644 --- a/include/linux/dma-buf.h +++ b/include/linux/dma-buf.h @@ -61,6 +61,10 @@ struct dma_buf_attachment; * This Callback must not sleep. * @kmap: maps a page from the buffer into kernel address space. * @kunmap: [optional] unmaps a page from the buffer. + * @mmap: used to expose the backing storage to userspace. Note that the + * mapping needs to be coherent - if the exporter doesn't directly + * support this, it needs to fake coherency by shooting down any ptes + * when transitioning away from the cpu domain. */ struct dma_buf_ops { int (*attach)(struct dma_buf *, struct device *, @@ -92,6 +96,8 @@ struct dma_buf_ops { void (*kunmap_atomic)(struct dma_buf *, unsigned long, void *); void *(*kmap)(struct dma_buf *, unsigned long); void (*kunmap)(struct dma_buf *, unsigned long, void *); + + int (*mmap)(struct dma_buf *, struct vm_area_struct *vma); }; /** @@ -167,6 +173,9 @@ void *dma_buf_kmap_atomic(struct dma_buf *, unsigned long); void dma_buf_kunmap_atomic(struct dma_buf *, unsigned long, void *); void *dma_buf_kmap(struct dma_buf *, unsigned long); void dma_buf_kunmap(struct dma_buf *, unsigned long, void *); + +int dma_buf_mmap(struct dma_buf *, struct vm_area_struct *, + unsigned long); #else static inline struct dma_buf_attachment *dma_buf_attach(struct dma_buf *dmabuf, @@ -248,6 +257,13 @@ static inline void dma_buf_kunmap(struct dma_buf *dmabuf, unsigned long pnum, void *vaddr) { } + +static inline int dma_buf_mmap(struct dma_buf *dmabuf, + struct vm_area_struct *vma, + unsigned long pgoff) +{ + return -ENODEV; +} #endif /* CONFIG_DMA_SHARED_BUFFER */ #endif /* __DMA_BUF_H__ */ -- 1.7.10

13 years, 4 months

[PATCH] dma-buf: add vmap interface (v2)

by Dave Airlie

From: Dave Airlie <airlied(a)redhat.com> The main requirement I have for this interface is for scanning out using the USB gpu devices. Since these devices have to read the framebuffer on updates and linearly compress it, using kmaps is a major overhead for every update. v2: fix warn issues pointed out by Sylwester Nawrocki. Signed-off-by: Dave Airlie <airlied(a)redhat.com> --- drivers/base/dma-buf.c | 34 ++++++++++++++++++++++++++++++++++ include/linux/dma-buf.h | 14 ++++++++++++++ 2 files changed, 48 insertions(+), 0 deletions(-) diff --git a/drivers/base/dma-buf.c b/drivers/base/dma-buf.c index 07cbbc6..750f92c 100644 --- a/drivers/base/dma-buf.c +++ b/drivers/base/dma-buf.c @@ -406,3 +406,37 @@ void dma_buf_kunmap(struct dma_buf *dmabuf, unsigned long page_num, dmabuf->ops->kunmap(dmabuf, page_num, vaddr); } EXPORT_SYMBOL_GPL(dma_buf_kunmap); + +/** + * dma_buf_vmap - Create virtual mapping for the buffer object into kernel address space. The same restrictions as for vmap and friends apply. + * @dma_buf: [in] buffer to vmap + * + * This call may fail due to lack of virtual mapping address space. + * These calls are optional in drivers. The intended use for them + * is for mapping objects linear in kernel space for high use objects. + * Please attempt to use kmap/kunmap before thinking about these interfaces. + */ +void *dma_buf_vmap(struct dma_buf *dmabuf) +{ + if (WARN_ON(!dmabuf)) + return NULL; + + if (dmabuf->ops->vmap) + return dmabuf->ops->vmap(dmabuf); + return NULL; +} +EXPORT_SYMBOL(dma_buf_vmap); + +/** + * dma_buf_vunmap - Unmap a vmap obtained by dma_buf_vmap. + * @dma_buf: [in] buffer to vmap + */ +void dma_buf_vunmap(struct dma_buf *dmabuf, void *vaddr) +{ + if (WARN_ON(!dmabuf)) + return; + + if (dmabuf->ops->vunmap) + dmabuf->ops->vunmap(dmabuf, vaddr); +} +EXPORT_SYMBOL(dma_buf_vunmap); diff --git a/include/linux/dma-buf.h b/include/linux/dma-buf.h index 3efbfc2..b92b6de 100644 --- a/include/linux/dma-buf.h +++ b/include/linux/dma-buf.h @@ -92,6 +92,9 @@ struct dma_buf_ops { void (*kunmap_atomic)(struct dma_buf *, unsigned long, void *); void *(*kmap)(struct dma_buf *, unsigned long); void (*kunmap)(struct dma_buf *, unsigned long, void *); + + void *(*vmap)(struct dma_buf *); + void (*vunmap)(struct dma_buf *, void *vaddr); }; /** @@ -167,6 +170,9 @@ void *dma_buf_kmap_atomic(struct dma_buf *, unsigned long); void dma_buf_kunmap_atomic(struct dma_buf *, unsigned long, void *); void *dma_buf_kmap(struct dma_buf *, unsigned long); void dma_buf_kunmap(struct dma_buf *, unsigned long, void *); + +void *dma_buf_vmap(struct dma_buf *); +void dma_buf_vunmap(struct dma_buf *, void *vaddr); #else static inline struct dma_buf_attachment *dma_buf_attach(struct dma_buf *dmabuf, @@ -248,6 +254,14 @@ static inline void dma_buf_kunmap(struct dma_buf *dmabuf, unsigned long pnum, void *vaddr) { } + +static inline void *dma_buf_vmap(struct dma_buf *) +{ +} + +static inline void dma_buf_vunmap(struct dma_buf *, void *vaddr); +{ +} #endif /* CONFIG_DMA_SHARED_BUFFER */ #endif /* __DMA_BUF_H__ */ -- 1.7.6

13 years, 4 months

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

Linaro-mm-sig