Linaro-mm-sig

linaro-mm-sig@lists.linaro.org

6 participants
3122 discussions

Re: [Linaro-mm-sig] [PATCH v2 0/3] drm/prime: Only call dma_map_sgtable() for devices with DMA support

by Daniel Vetter

On Mon, Feb 22, 2021 at 2:24 PM Thomas Zimmermann <tzimmermann(a)suse.de> wrote: > > Hi > > Am 22.02.21 um 14:09 schrieb Christian König: > > > > > > Am 22.02.21 um 13:43 schrieb Thomas Zimmermann: > >> USB-based drivers cannot use DMA, so the importing of dma-buf attachments > >> currently fails for udl and gm12u320. This breaks joining/mirroring of > >> displays. > >> > >> The fix is now a little series. To solve the issue on the importer > >> side (i.e., the affected USB-based driver), patch 1 introduces a new > >> PRIME callback, struct drm_driver.gem_prime_create_object, which creates > >> an object and gives more control to the importing driver. Specifically, > >> udl and gm12u320 can now avoid the creation of a scatter/gather table > >> for the imported pages. Patch 1 is self-contained in the sense that it > >> can be backported into older kernels. > > > > Mhm, that sounds like a little overkill to me. > > > > Drivers can already import the DMA-bufs all by them selves without the > > help of the DRM functions. See amdgpu for an example. > > > > Daniel also already noted to me that he sees the DRM helper as a bit > > questionable middle layer. > > And this bug proves that it is. :) The trouble here is actually gem_bo->import_attach, which isn't really part of the questionable midlayer, but fairly mandatory (only exception is vmwgfx because not using gem) caching to make sure we don't end up with duped imports and fun stuff like that. And dma_buf_attach now implicitly creates the sg table already, so we're already in game over land. I think we'd need to make import_attach a union with import_buf or something like that, so that you can do attachment-less importing. > > Have you thought about doing that instead? > > There appears to be some useful code in drm_gem_prime_import_dev(). But > if the general sentiment goes towards removing > gem_prime_import_sg_table, we can work towards that as well. I still think this part is a bit a silly midlayer for no good reason, but I think that's orthogonal to the issue at hand here. I'd suggest we first try to paper over the issue by using prime_import_dev with the host controller (which hopefully is dma-capable for most systems). And then, at leisure, try to untangle the obj->import_attach issue. -Daniel > > Best regards > Thomas > > > > > Christian. > > > >> > >> Patches 2 and 3 update SHMEM and CMA helpers to use the new callback. > >> Effectively this moves the sg table setup from the PRIME helpers into > >> the memory managers. SHMEM now supports devices without DMA support, > >> so custom code can be removed from udl and g12u320. > >> > >> Tested by joining/mirroring displays of udl and radeon under Gnome/X11. > >> > >> v2: > >> * move fix to importer side (Christian, Daniel) > >> * update SHMEM and CMA helpers for new PRIME callbacks > >> > >> Thomas Zimmermann (3): > >> drm: Support importing dmabufs into drivers without DMA > >> drm/shmem-helper: Implement struct drm_driver.gem_prime_create_object > >> drm/cma-helper: Implement struct drm_driver.gem_prime_create_object > >> > >> drivers/gpu/drm/drm_gem_cma_helper.c | 62 ++++++++++++++----------- > >> drivers/gpu/drm/drm_gem_shmem_helper.c | 38 ++++++++++----- > >> drivers/gpu/drm/drm_prime.c | 43 +++++++++++------ > >> drivers/gpu/drm/lima/lima_drv.c | 2 +- > >> drivers/gpu/drm/panfrost/panfrost_drv.c | 2 +- > >> drivers/gpu/drm/panfrost/panfrost_gem.c | 6 +-- > >> drivers/gpu/drm/panfrost/panfrost_gem.h | 4 +- > >> drivers/gpu/drm/pl111/pl111_drv.c | 8 ++-- > >> drivers/gpu/drm/v3d/v3d_bo.c | 6 +-- > >> drivers/gpu/drm/v3d/v3d_drv.c | 2 +- > >> drivers/gpu/drm/v3d/v3d_drv.h | 5 +- > >> include/drm/drm_drv.h | 12 +++++ > >> include/drm/drm_gem_cma_helper.h | 12 ++--- > >> include/drm/drm_gem_shmem_helper.h | 6 +-- > >> 14 files changed, 120 insertions(+), 88 deletions(-) > >> > >> -- > >> 2.30.1 > >> > > > > -- > Thomas Zimmermann > Graphics Driver Developer > SUSE Software Solutions Germany GmbH > Maxfeldstr. 5, 90409 Nürnberg, Germany > (HRB 36809, AG Nürnberg) > Geschäftsführer: Felix Imendörffer > -- Daniel Vetter Software Engineer, Intel Corporation http://blog.ffwll.ch

4 years, 8 months

Re: [Linaro-mm-sig] [PATCH v2 0/3] drm/prime: Only call dma_map_sgtable() for devices with DMA support

by Christian König

Am 22.02.21 um 13:43 schrieb Thomas Zimmermann: > USB-based drivers cannot use DMA, so the importing of dma-buf attachments > currently fails for udl and gm12u320. This breaks joining/mirroring of > displays. > > The fix is now a little series. To solve the issue on the importer > side (i.e., the affected USB-based driver), patch 1 introduces a new > PRIME callback, struct drm_driver.gem_prime_create_object, which creates > an object and gives more control to the importing driver. Specifically, > udl and gm12u320 can now avoid the creation of a scatter/gather table > for the imported pages. Patch 1 is self-contained in the sense that it > can be backported into older kernels. Mhm, that sounds like a little overkill to me. Drivers can already import the DMA-bufs all by them selves without the help of the DRM functions. See amdgpu for an example. Daniel also already noted to me that he sees the DRM helper as a bit questionable middle layer. Have you thought about doing that instead? Christian. > > Patches 2 and 3 update SHMEM and CMA helpers to use the new callback. > Effectively this moves the sg table setup from the PRIME helpers into > the memory managers. SHMEM now supports devices without DMA support, > so custom code can be removed from udl and g12u320. > > Tested by joining/mirroring displays of udl and radeon under Gnome/X11. > > v2: > * move fix to importer side (Christian, Daniel) > * update SHMEM and CMA helpers for new PRIME callbacks > > Thomas Zimmermann (3): > drm: Support importing dmabufs into drivers without DMA > drm/shmem-helper: Implement struct drm_driver.gem_prime_create_object > drm/cma-helper: Implement struct drm_driver.gem_prime_create_object > > drivers/gpu/drm/drm_gem_cma_helper.c | 62 ++++++++++++++----------- > drivers/gpu/drm/drm_gem_shmem_helper.c | 38 ++++++++++----- > drivers/gpu/drm/drm_prime.c | 43 +++++++++++------ > drivers/gpu/drm/lima/lima_drv.c | 2 +- > drivers/gpu/drm/panfrost/panfrost_drv.c | 2 +- > drivers/gpu/drm/panfrost/panfrost_gem.c | 6 +-- > drivers/gpu/drm/panfrost/panfrost_gem.h | 4 +- > drivers/gpu/drm/pl111/pl111_drv.c | 8 ++-- > drivers/gpu/drm/v3d/v3d_bo.c | 6 +-- > drivers/gpu/drm/v3d/v3d_drv.c | 2 +- > drivers/gpu/drm/v3d/v3d_drv.h | 5 +- > include/drm/drm_drv.h | 12 +++++ > include/drm/drm_gem_cma_helper.h | 12 ++--- > include/drm/drm_gem_shmem_helper.h | 6 +-- > 14 files changed, 120 insertions(+), 88 deletions(-) > > -- > 2.30.1 >

4 years, 8 months

[PATCH] drm-buf: Add debug option

by Daniel Vetter

We have too many people abusing the struct page they can get at but really shouldn't in importers. Aside from that the backing page might simply not exist (for dynamic p2p mappings) looking at it and using it e.g. for mmap can also wreak the page handling of the exporter completely. Importers really must go through the proper interface like dma_buf_mmap for everything. Just an RFC to see whether this idea has some stickiness. default y for now to make sure intel-gfx-ci picks it up too. I'm semi-tempted to enforce this for dynamic importers since those really have no excuse at all to break the rules. Unfortuantely we can't store the right pointers somewhere safe to make sure we oops on something recognizable, so best is to just wrangle them a bit by flipping all the bits. At least on x86 kernel addresses have all their high bits sets and the struct page array is fairly low in the kernel mapping, so flipping all the bits gives us a very high pointer in userspace and hence excellent chances for an invalid dereference. Signed-off-by: Daniel Vetter <daniel.vetter(a)intel.com> Cc: Sumit Semwal <sumit.semwal(a)linaro.org> Cc: "Christian König" <christian.koenig(a)amd.com> Cc: David Stevens <stevensd(a)chromium.org> Cc: linux-media(a)vger.kernel.org Cc: linaro-mm-sig(a)lists.linaro.org --- drivers/dma-buf/Kconfig | 8 +++++++ drivers/dma-buf/dma-buf.c | 49 +++++++++++++++++++++++++++++++++++---- 2 files changed, 53 insertions(+), 4 deletions(-) diff --git a/drivers/dma-buf/Kconfig b/drivers/dma-buf/Kconfig index 4f8224a6ac95..cddb549e5e59 100644 --- a/drivers/dma-buf/Kconfig +++ b/drivers/dma-buf/Kconfig @@ -50,6 +50,14 @@ config DMABUF_MOVE_NOTIFY This is marked experimental because we don't yet have a consistent execution context and memory management between drivers. +config DMABUF_DEBUG + bool "DMA-BUF debug checks" + default y + help + This option enables additional checks for DMA-BUF importers and + exporters. Specifically it validates that importers do not peek at the + underlying struct page when they import a buffer. + config DMABUF_SELFTESTS tristate "Selftests for the dma-buf interfaces" default n diff --git a/drivers/dma-buf/dma-buf.c b/drivers/dma-buf/dma-buf.c index 1c9bd51db110..6e4725f7dfde 100644 --- a/drivers/dma-buf/dma-buf.c +++ b/drivers/dma-buf/dma-buf.c @@ -666,6 +666,30 @@ void dma_buf_put(struct dma_buf *dmabuf) } EXPORT_SYMBOL_GPL(dma_buf_put); +static struct sg_table * __map_dma_buf(struct dma_buf_attachment *attach, + enum dma_data_direction direction) +{ + struct sg_table *sg_table; + + sg_table = attach->dmabuf->ops->map_dma_buf(attach, direction); + +#if CONFIG_DMABUF_DEBUG + if (sg_table) { + int i; + struct scatterlist *sg; + + /* To catch abuse of the underlying struct page by importers mix + * up the bits, but take care to preserve the low SG_ bits to + * not corrupt the sgt. The mixing is undone in __unmap_dma_buf + * before passing the sgt back to the exporter. */ + for_each_sgtable_sg(sg_table, sg, i) + sg->page_link ^= ~0xffUL; + } +#endif + + return sg_table; +} + /** * dma_buf_dynamic_attach - Add the device to dma_buf's attachments list * @dmabuf: [in] buffer to attach device to. @@ -737,7 +761,7 @@ dma_buf_dynamic_attach(struct dma_buf *dmabuf, struct device *dev, goto err_unlock; } - sgt = dmabuf->ops->map_dma_buf(attach, DMA_BIDIRECTIONAL); + sgt = __map_dma_buf(attach, DMA_BIDIRECTIONAL); if (!sgt) sgt = ERR_PTR(-ENOMEM); if (IS_ERR(sgt)) { @@ -784,6 +808,23 @@ struct dma_buf_attachment *dma_buf_attach(struct dma_buf *dmabuf, } EXPORT_SYMBOL_GPL(dma_buf_attach); +static void __unmap_dma_buf(struct dma_buf_attachment *attach, + struct sg_table *sg_table, + enum dma_data_direction direction) +{ + +#if CONFIG_DMABUF_DEBUG + if (sg_table) { + int i; + struct scatterlist *sg; + + for_each_sgtable_sg(sg_table, sg, i) + sg->page_link ^= ~0xffUL; + } +#endif + attach->dmabuf->ops->unmap_dma_buf(attach, sg_table, direction); +} + /** * dma_buf_detach - Remove the given attachment from dmabuf's attachments list * @dmabuf: [in] buffer to detach from. @@ -802,7 +843,7 @@ void dma_buf_detach(struct dma_buf *dmabuf, struct dma_buf_attachment *attach) if (dma_buf_is_dynamic(attach->dmabuf)) dma_resv_lock(attach->dmabuf->resv, NULL); - dmabuf->ops->unmap_dma_buf(attach, attach->sgt, attach->dir); + __unmap_dma_buf(attach, attach->sgt, attach->dir); if (dma_buf_is_dynamic(attach->dmabuf)) { dma_buf_unpin(attach); @@ -924,7 +965,7 @@ struct sg_table *dma_buf_map_attachment(struct dma_buf_attachment *attach, } } - sg_table = attach->dmabuf->ops->map_dma_buf(attach, direction); + sg_table = __map_dma_buf(attach, direction); if (!sg_table) sg_table = ERR_PTR(-ENOMEM); @@ -987,7 +1028,7 @@ void dma_buf_unmap_attachment(struct dma_buf_attachment *attach, if (dma_buf_is_dynamic(attach->dmabuf)) dma_resv_assert_held(attach->dmabuf->resv); - attach->dmabuf->ops->unmap_dma_buf(attach, sg_table, direction); + __unmap_dma_buf(attach, sg_table, direction); if (dma_buf_is_dynamic(attach->dmabuf) && !IS_ENABLED(CONFIG_DMABUF_MOVE_NOTIFY)) -- 2.29.2

4 years, 8 months

Re: [Linaro-mm-sig] DMA-buf and uncached system memory

by Christian König

Am 15.02.21 um 10:06 schrieb Simon Ser: > On Monday, February 15th, 2021 at 9:58 AM, Christian König <christian.koenig(a)amd.com> wrote: > >> we are currently working an Freesync and direct scan out from system >> memory on AMD APUs in A+A laptops. >> >> On problem we stumbled over is that our display hardware needs to scan >> out from uncached system memory and we currently don't have a way to >> communicate that through DMA-buf. >> >> For our specific use case at hand we are going to implement something >> driver specific, but the question is should we have something more >> generic for this? >> >> After all the system memory access pattern is a PCIe extension and as >> such something generic. > Intel also needs uncached system memory if I'm not mistaken? No idea, that's why I'm asking. Could be that this is also interesting for I+A systems. > Where are the buffers allocated? If GBM, then it needs to allocate memory that > can be scanned out if the USE_SCANOUT flag is set or if a scanout-capable > modifier is picked. > > If this is about communicating buffer constraints between different components > of the stack, there were a few proposals about it. The most recent one is [1]. Well the problem here is on a different level of the stack. See resolution, pitch etc:.. can easily communicated in userspace without involvement of the kernel. The worst thing which can happen is that you draw garbage into your own application window. But if you get the caching attributes in the page tables (both CPU as well as IOMMU, device etc...) wrong then ARM for example has the tendency to just spontaneously reboot X86 is fortunately a bit more gracefully and you only end up with random data corruption, but that is only marginally better. So to sum it up that is not something which we can leave in the hands of userspace. I think that exporters in the DMA-buf framework should have the ability to tell importers if the system memory snooping is necessary or not. Userspace components can then of course tell the exporter what the importer needs, but validation if that stuff is correct and doesn't crash the system must happen in the kernel. Regards, Christian. > > Simon > > [1]: https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fxdc2020.x…

4 years, 8 months

Re: [Linaro-mm-sig] DMA-buf and uncached system memory

by Christian König

Am 15.02.21 um 13:00 schrieb Thomas Zimmermann: > Hi > > Am 15.02.21 um 10:49 schrieb Thomas Zimmermann: >> Hi >> >> Am 15.02.21 um 09:58 schrieb Christian König: >>> Hi guys, >>> >>> we are currently working an Freesync and direct scan out from system >>> memory on AMD APUs in A+A laptops. >>> >>> On problem we stumbled over is that our display hardware needs to >>> scan out from uncached system memory and we currently don't have a >>> way to communicate that through DMA-buf. > > Re-reading this paragrah, it sounds more as if you want to let the > exporter know where to move the buffer. Is this another case of the > missing-pin-flag problem? No, your original interpretation was correct. Maybe my writing is a bit unspecific. The real underlying issue is that our display hardware has a problem with latency when accessing system memory. So the question is if that also applies to for example Intel hardware or other devices as well or if it is just something AMD specific? Regards, Christian. > > Best regards > Thomas > >>> >>> For our specific use case at hand we are going to implement >>> something driver specific, but the question is should we have >>> something more generic for this? >> >> For vmap operations, we return the address as struct dma_buf_map, >> which contains additional information about the memory buffer. In >> vram helpers, we have the interface drm_gem_vram_offset() that >> returns the offset of the GPU device memory. >> >> Would it be feasible to combine both concepts into a dma-buf >> interface that returns the device-memory offset plus the additional >> caching flag? >> >> There'd be a structure and a getter function returning the structure. >> >> struct dma_buf_offset { >> bool cached; >> u64 address; >> }; >> >> // return offset in *off >> int dma_buf_offset(struct dma_buf *buf, struct dma_buf_off *off); >> >> Whatever settings are returned by dma_buf_offset() are valid while >> the dma_buf is pinned. >> >> Best regards >> Thomas >> >>> >>> After all the system memory access pattern is a PCIe extension and >>> as such something generic. >>> >>> Regards, >>> Christian. >>> _______________________________________________ >>> dri-devel mailing list >>> dri-devel(a)lists.freedesktop.org >>> https://lists.freedesktop.org/mailman/listinfo/dri-devel >> >> >> _______________________________________________ >> dri-devel mailing list >> dri-devel(a)lists.freedesktop.org >> https://lists.freedesktop.org/mailman/listinfo/dri-devel >> >

4 years, 8 months

Re: [Linaro-mm-sig] [PATCH] RFC: dma-fence: Document recoverable page fault implications

by Daniel Vetter

On Tue, Feb 9, 2021 at 4:13 AM Bas Nieuwenhuizen <bas(a)basnieuwenhuizen.nl> wrote: > > On Thu, Jan 28, 2021 at 4:40 PM Felix Kuehling <felix.kuehling(a)amd.com> wrote: > > > > Am 2021-01-28 um 2:39 a.m. schrieb Christian König: > > > Am 27.01.21 um 23:00 schrieb Felix Kuehling: > > >> Am 2021-01-27 um 7:16 a.m. schrieb Christian König: > > >>> Am 27.01.21 um 13:11 schrieb Maarten Lankhorst: > > >>>> Op 27-01-2021 om 01:22 schreef Felix Kuehling: > > >>>>> Am 2021-01-21 um 2:40 p.m. schrieb Daniel Vetter: > > >>>>>> Recently there was a fairly long thread about recoreable hardware > > >>>>>> page > > >>>>>> faults, how they can deadlock, and what to do about that. > > >>>>>> > > >>>>>> While the discussion is still fresh I figured good time to try and > > >>>>>> document the conclusions a bit. > > >>>>>> > > >>>>>> References: > > >>>>>> https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Flore.kern… > > >>>>>> > > >>>>>> Cc: Maarten Lankhorst <maarten.lankhorst(a)linux.intel.com> > > >>>>>> Cc: Thomas Hellström <thomas.hellstrom(a)intel.com> > > >>>>>> Cc: "Christian König" <christian.koenig(a)amd.com> > > >>>>>> Cc: Jerome Glisse <jglisse(a)redhat.com> > > >>>>>> Cc: Felix Kuehling <felix.kuehling(a)amd.com> > > >>>>>> Signed-off-by: Daniel Vetter <daniel.vetter(a)intel.com> > > >>>>>> Cc: Sumit Semwal <sumit.semwal(a)linaro.org> > > >>>>>> Cc: linux-media(a)vger.kernel.org > > >>>>>> Cc: linaro-mm-sig(a)lists.linaro.org > > >>>>>> -- > > >>>>>> I'll be away next week, but figured I'll type this up quickly for > > >>>>>> some > > >>>>>> comments and to check whether I got this all roughly right. > > >>>>>> > > >>>>>> Critique very much wanted on this, so that we can make sure hw which > > >>>>>> can't preempt (with pagefaults pending) like gfx10 has a clear > > >>>>>> path to > > >>>>>> support page faults in upstream. So anything I missed, got wrong or > > >>>>>> like that would be good. > > >>>>>> -Daniel > > >>>>>> --- > > >>>>>> Documentation/driver-api/dma-buf.rst | 66 > > >>>>>> ++++++++++++++++++++++++++++ > > >>>>>> 1 file changed, 66 insertions(+) > > >>>>>> > > >>>>>> diff --git a/Documentation/driver-api/dma-buf.rst > > >>>>>> b/Documentation/driver-api/dma-buf.rst > > >>>>>> index a2133d69872c..e924c1e4f7a3 100644 > > >>>>>> --- a/Documentation/driver-api/dma-buf.rst > > >>>>>> +++ b/Documentation/driver-api/dma-buf.rst > > >>>>>> @@ -257,3 +257,69 @@ fences in the kernel. This means: > > >>>>>> userspace is allowed to use userspace fencing or long running > > >>>>>> compute > > >>>>>> workloads. This also means no implicit fencing for shared > > >>>>>> buffers in these > > >>>>>> cases. > > >>>>>> + > > >>>>>> +Recoverable Hardware Page Faults Implications > > >>>>>> +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ > > >>>>>> + > > >>>>>> +Modern hardware supports recoverable page faults, which has a > > >>>>>> lot of > > >>>>>> +implications for DMA fences. > > >>>>>> + > > >>>>>> +First, a pending page fault obviously holds up the work that's > > >>>>>> running on the > > >>>>>> +accelerator and a memory allocation is usually required to resolve > > >>>>>> the fault. > > >>>>>> +But memory allocations are not allowed to gate completion of DMA > > >>>>>> fences, which > > >>>>>> +means any workload using recoverable page faults cannot use DMA > > >>>>>> fences for > > >>>>>> +synchronization. Synchronization fences controlled by userspace > > >>>>>> must be used > > >>>>>> +instead. > > >>>>>> + > > >>>>>> +On GPUs this poses a problem, because current desktop compositor > > >>>>>> protocols on > > >>>>>> +Linus rely on DMA fences, which means without an entirely new > > >>>>>> userspace stack > > >>>>>> +built on top of userspace fences, they cannot benefit from > > >>>>>> recoverable page > > >>>>>> +faults. The exception is when page faults are only used as > > >>>>>> migration hints and > > >>>>>> +never to on-demand fill a memory request. For now this means > > >>>>>> recoverable page > > >>>>>> +faults on GPUs are limited to pure compute workloads. > > >>>>>> + > > >>>>>> +Furthermore GPUs usually have shared resources between the 3D > > >>>>>> rendering and > > >>>>>> +compute side, like compute units or command submission engines. If > > >>>>>> both a 3D > > >>>>>> +job with a DMA fence and a compute workload using recoverable page > > >>>>>> faults are > > >>>>>> +pending they could deadlock: > > >>>>>> + > > >>>>>> +- The 3D workload might need to wait for the compute job to finish > > >>>>>> and release > > >>>>>> + hardware resources first. > > >>>>>> + > > >>>>>> +- The compute workload might be stuck in a page fault, because the > > >>>>>> memory > > >>>>>> + allocation is waiting for the DMA fence of the 3D workload to > > >>>>>> complete. > > >>>>>> + > > >>>>>> +There are a few ways to prevent this problem: > > >>>>>> + > > >>>>>> +- Compute workloads can always be preempted, even when a page > > >>>>>> fault is pending > > >>>>>> + and not yet repaired. Not all hardware supports this. > > >>>>>> + > > >>>>>> +- DMA fence workloads and workloads which need page fault handling > > >>>>>> have > > >>>>>> + independent hardware resources to guarantee forward progress. > > >>>>>> This could be > > >>>>>> + achieved through e.g. through dedicated engines and minimal > > >>>>>> compute unit > > >>>>>> + reservations for DMA fence workloads. > > >>>>>> + > > >>>>>> +- The reservation approach could be further refined by only > > >>>>>> reserving the > > >>>>>> + hardware resources for DMA fence workloads when they are > > >>>>>> in-flight. This must > > >>>>>> + cover the time from when the DMA fence is visible to other > > >>>>>> threads up to > > >>>>>> + moment when fence is completed through dma_fence_signal(). > > >>>>>> + > > >>>>>> +- As a last resort, if the hardware provides no useful reservation > > >>>>>> mechanics, > > >>>>>> + all workloads must be flushed from the GPU when switching > > >>>>>> between jobs > > >>>>>> + requiring DMA fences or jobs requiring page fault handling: This > > >>>>>> means all DMA > > >>>>>> + fences must complete before a compute job with page fault > > >>>>>> handling can be > > >>>>>> + inserted into the scheduler queue. And vice versa, before a DMA > > >>>>>> fence can be > > >>>>>> + made visible anywhere in the system, all compute workloads must > > >>>>>> be preempted > > >>>>>> + to guarantee all pending GPU page faults are flushed. > > >>>>> I thought of another possible workaround: > > >>>>> > > >>>>> * Partition the memory. Servicing of page faults will use a > > >>>>> separate > > >>>>> memory pool that can always be allocated from without > > >>>>> waiting for > > >>>>> fences. This includes memory for page tables and memory for > > >>>>> migrating data to. You may steal memory from other processes > > >>>>> that > > >>>>> can page fault, so no fence waiting is necessary. Being able to > > >>>>> steal memory at any time also means there are basically no > > >>>>> out-of-memory situations you need to worry about. Even page > > >>>>> tables > > >>>>> (except the root page directory of each process) can be > > >>>>> stolen in > > >>>>> the worst case. > > >>>> I think 'overcommit' would be a nice way to describe this. But I'm not > > >>>> sure how easy this is to implement in practice. You would basically > > >>>> need > > >>>> to create your own memory manager for this. > > >>> Well you would need a completely separate pool for both device as well > > >>> as system memory. > > >>> > > >>> E.g. on boot we say we steal X GB system memory only for HMM. > > >> Why? The GPU driver doesn't need to allocate system memory for HMM. > > >> Migrations to system memory are handled by the kernel's handle_mm_fault > > >> and page allocator and swap logic. > > > > > > And that one depends on dma_fence completion because you can easily > > > need to wait for an MMU notifier callback. > > > > I see, the GFX MMU notifier for userpointers in amdgpu currently waits > > for fences. For the KFD MMU notifier I am planning to fix this by > > causing GPU page faults instead of preempting the queues. Can we limit > > userptrs in amdgpu to engines that can page fault. Basically make it > > illegal to attach userptr BOs to graphics CS BO lists, so they can only > > be used in user mode command submissions, which can page fault. Then the > > GFX MMU notifier could invalidate PTEs and would not have to wait for > > fences. > > sadly graphics + userptr is already exposed via Mesa. This is not about userptr, we fake userptr entirely in software. It's about exposing recoverable gpu page faults (which would make userptr maybe more efficient since we could do on-demand paging). userptr itself isn't a problem, but it is part of the reasons why this is tricky. Christian/Felix, I think for kernel folks this is clear enough that I don't need to clarify this in the text? -Daniel > > > > > > > > > > > As Maarten wrote when you want to go down this route you need a > > > complete separate memory management parallel to the one of the kernel. > > > > Not really. I'm trying to make the GPU memory management more similar to > > what the kernel does for system memory. > > > > I understood Maarten's comment as "I'm creating a new memory manager and > > not using TTM any more". This is true. The idea is that this portion of > > VRAM would be managed more like system memory. > > > > Regards, > > Felix > > > > > > > > > > Regards, > > > Christian. > > > > > >> It doesn't depend on any fences, so > > >> it cannot deadlock with any GPU driver-managed memory. The GPU driver > > >> gets involved in the MMU notifier to invalidate device page tables. But > > >> that also doesn't need to wait for any fences. > > >> > > >> And if the kernel runs out of pageable memory, you're in trouble anyway. > > >> The OOM killer will step in, nothing new there. > > >> > > >> Regards, > > >> Felix > > >> > > >> > > >>>> But from a design point of view, definitely a valid solution. > > >>> I think the restriction above makes it pretty much unusable. > > >>> > > >>>> But this looks good, those solutions are definitely the valid > > >>>> options we > > >>>> can choose from. > > >>> It's certainly worth noting, yes. And just to make sure that nobody > > >>> has the idea to reserve only device memory. > > >>> > > >>> Christian. > > >>> > > >>>> ~Maarten > > >>>> > > >> _______________________________________________ > > >> Linaro-mm-sig mailing list > > >> Linaro-mm-sig(a)lists.linaro.org > > >> https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Flists.lin… > > >> > > > > > _______________________________________________ > > dri-devel mailing list > > dri-devel(a)lists.freedesktop.org > > https://lists.freedesktop.org/mailman/listinfo/dri-devel -- Daniel Vetter Software Engineer, Intel Corporation http://blog.ffwll.ch

4 years, 8 months

[PATCH v4 0/4] Chunk Heap Support on DMA-HEAP

by Minchan Kim

This patchset introduces a new dma heap, "chunk-heap" that makes it easy to perform the bulk allocation of high order pages. It has been created to help optimize the 4K/8K HDR video playback with secure DRM HW to protect contents on memory. The HW needs physically contiguous memory chunks(e.g, 64K) up to several hundred MB memory. To make such high-order big bulk allocations work, chunk-heap uses CMA area. To avoid CMA allocation long stall on blocking pages(e.g., page writeback and/or page locking), it uses failfast mode of the CMA API(i.e., __GFP_NORETRY) so it will continue to find easy migratable pages in different pageblocks without stalling. At last resort, it will allow the blocking only if it couldn't find the available memory in the end. First two patches introduces the failfast mode as __GFP_NORETRY in alloc_contig_range and the allow to use it from the CMA API. Third patch introduces device tree syntax for chunk-heap to bind the specific CMA area with chunk-heap. Finally, last patch implements chunk-heap as dma-buf heap. * since v3 - https://lore.kernel.org/linux-mm/20210113012143.1201105-1-minchan@kernel.or… * use prefix for chunk-name - John * fix yamllint error - Rob * add reviewed-by - Suren * since v2 - https://lore.kernel.org/linux-mm/20201201175144.3996569-1-minchan@kernel.or… * introduce gfp_mask with __GFP_NORETRY on cma_alloc - Michal * do not expoert CMA APIs - Christoph * use compatible string for DT instead of dma-heap specific property - Hridya * Since v1 - https://lore.kernel.org/linux-mm/20201117181935.3613581-1-minchan@kernel.or… * introduce alloc_contig_mode - David * use default CMA instead of device tree - John Hyesoo Yu (2): dt-bindings: reserved-memory: Make DMA-BUF CMA heap DT-configurable dma-buf: heaps: add chunk heap to dmabuf heaps Minchan Kim (2): mm: cma: introduce gfp flag in cma_alloc instead of no_warn mm: failfast mode with __GFP_NORETRY in alloc_contig_range .../reserved-memory/dma_heap_chunk.yaml | 56 ++ drivers/dma-buf/heaps/Kconfig | 8 + drivers/dma-buf/heaps/Makefile | 1 + drivers/dma-buf/heaps/chunk_heap.c | 492 ++++++++++++++++++ drivers/dma-buf/heaps/cma_heap.c | 2 +- drivers/s390/char/vmcp.c | 2 +- include/linux/cma.h | 2 +- kernel/dma/contiguous.c | 3 +- mm/cma.c | 12 +- mm/cma_debug.c | 2 +- mm/hugetlb.c | 6 +- mm/page_alloc.c | 8 +- mm/secretmem.c | 3 +- 13 files changed, 581 insertions(+), 16 deletions(-) create mode 100644 Documentation/devicetree/bindings/reserved-memory/dma_heap_chunk.yaml create mode 100644 drivers/dma-buf/heaps/chunk_heap.c -- 2.30.0.296.g2bfb1c46d8-goog

4 years, 9 months

Re: [Linaro-mm-sig] [PATCH v3 2/2] dmabuf: Add dmabuf inode number to /proc/*/fdinfo

by Christian König

Am 05.02.21 um 03:23 schrieb Kalesh Singh: > If a FD refers to a DMA buffer add the DMA buffer inode number to > /proc/<pid>/fdinfo/<FD> and /proc/<pid>/task/<tid>/fdindo/<FD>. > > The dmabuf inode number allows userspace to uniquely identify the buffer > and avoids a dependency on /proc/<pid>/fd/* when accounting per-process > DMA buffer sizes. BTW: Why do we make this DMA-buf specific? Couldn't we always print the inode number for all fds? Regards, Christian. > > Signed-off-by: Kalesh Singh <kaleshsingh(a)google.com> > --- > Changes in v3: > - Add documentation in proc.rst > Changes in v2: > - Update patch description > > Documentation/filesystems/proc.rst | 17 +++++++++++++++++ > drivers/dma-buf/dma-buf.c | 1 + > 2 files changed, 18 insertions(+) > > diff --git a/Documentation/filesystems/proc.rst b/Documentation/filesystems/proc.rst > index 2fa69f710e2a..fdd38676f57f 100644 > --- a/Documentation/filesystems/proc.rst > +++ b/Documentation/filesystems/proc.rst > @@ -2031,6 +2031,23 @@ details]. 'it_value' is remaining time until the timer expiration. > with TIMER_ABSTIME option which will be shown in 'settime flags', but 'it_value' > still exhibits timer's remaining time. > > +DMA Buffer files > +~~~~~~~~~~~~~~~~ > + > +:: > + > + pos: 0 > + flags: 04002 > + mnt_id: 9 > + dmabuf_inode_no: 63107 > + size: 32768 > + count: 2 > + exp_name: system-heap > + > +where 'dmabuf_inode_no' is the unique inode number of the DMA buffer file. > +'size' is the size of the DMA buffer in bytes. 'count' is the file count of > +the DMA buffer file. 'exp_name' is the name of the DMA buffer exporter. > + > 3.9 /proc/<pid>/map_files - Information about memory mapped files > --------------------------------------------------------------------- > This directory contains symbolic links which represent memory mapped files > diff --git a/drivers/dma-buf/dma-buf.c b/drivers/dma-buf/dma-buf.c > index 9ad6397aaa97..d869099ede83 100644 > --- a/drivers/dma-buf/dma-buf.c > +++ b/drivers/dma-buf/dma-buf.c > @@ -414,6 +414,7 @@ static void dma_buf_show_fdinfo(struct seq_file *m, struct file *file) > { > struct dma_buf *dmabuf = file->private_data; > > + seq_printf(m, "dmabuf_inode_no:\t%lu\n", file_inode(file)->i_ino); > seq_printf(m, "size:\t%zu\n", dmabuf->size); > /* Don't count the temporary reference taken inside procfs seq_show */ > seq_printf(m, "count:\t%ld\n", file_count(dmabuf->file) - 1);

4 years, 9 months

[PATCH] RFC: dma-buf: Require VM_SPECIAL vma for mmap

by Daniel Vetter

tldr; DMA buffers aren't normal memory, expecting that you can use them like that (like calling get_user_pages works, or that they're accounting like any other normal memory) cannot be guaranteed. Since some userspace only runs on integrated devices, where all buffers are actually all resident system memory, there's a huge temptation to assume that a struct page is always present and useable like for any more pagecache backed mmap. This has the potential to result in a uapi nightmare. To stop this gap require that DMA buffer mmaps are VM_SPECIAL, which blocks get_user_pages and all the other struct page based infrastructure for everyone. In spirit this is the uapi counterpart to the kernel-internal CONFIG_DMABUF_DEBUG. Motivated by a recent patch which wanted to swich the system dma-buf heap to vm_insert_page instead of vm_insert_pfn. References: https://lore.kernel.org/lkml/CAKMK7uHi+mG0z0HUmNt13QCCvutuRVjpcR0NjRL12k-Wb… Cc: Jason Gunthorpe <jgg(a)ziepe.ca> Cc: Suren Baghdasaryan <surenb(a)google.com> Cc: Matthew Wilcox <willy(a)infradead.org> Cc: John Stultz <john.stultz(a)linaro.org> Signed-off-by: Daniel Vetter <daniel.vetter(a)intel.com> Cc: Sumit Semwal <sumit.semwal(a)linaro.org> Cc: "Christian König" <christian.koenig(a)amd.com> Cc: linux-media(a)vger.kernel.org Cc: linaro-mm-sig(a)lists.linaro.org --- drivers/dma-buf/dma-buf.c | 15 +++++++++++++-- 1 file changed, 13 insertions(+), 2 deletions(-) diff --git a/drivers/dma-buf/dma-buf.c b/drivers/dma-buf/dma-buf.c index f264b70c383e..d3081fc07056 100644 --- a/drivers/dma-buf/dma-buf.c +++ b/drivers/dma-buf/dma-buf.c @@ -127,6 +127,7 @@ static struct file_system_type dma_buf_fs_type = { static int dma_buf_mmap_internal(struct file *file, struct vm_area_struct *vma) { struct dma_buf *dmabuf; + int ret; if (!is_dma_buf_file(file)) return -EINVAL; @@ -142,7 +143,11 @@ static int dma_buf_mmap_internal(struct file *file, struct vm_area_struct *vma) dmabuf->size >> PAGE_SHIFT) return -EINVAL; - return dmabuf->ops->mmap(dmabuf, vma); + ret = dmabuf->ops->mmap(dmabuf, vma); + + WARN_ON(!(vma->vm_flags & VM_SPECIAL)); + + return ret; } static loff_t dma_buf_llseek(struct file *file, loff_t offset, int whence) @@ -1244,6 +1249,8 @@ EXPORT_SYMBOL_GPL(dma_buf_end_cpu_access); int dma_buf_mmap(struct dma_buf *dmabuf, struct vm_area_struct *vma, unsigned long pgoff) { + int ret; + if (WARN_ON(!dmabuf || !vma)) return -EINVAL; @@ -1264,7 +1271,11 @@ int dma_buf_mmap(struct dma_buf *dmabuf, struct vm_area_struct *vma, vma_set_file(vma, dmabuf->file); vma->vm_pgoff = pgoff; - return dmabuf->ops->mmap(dmabuf, vma); + ret = dmabuf->ops->mmap(dmabuf, vma); + + WARN_ON(!(vma->vm_flags & VM_SPECIAL)); + + return ret; } EXPORT_SYMBOL_GPL(dma_buf_mmap); -- 2.30.0

4 years, 9 months

[PATCH 1/2] mm: replace BUG_ON in vm_insert_page with a return of an error

by Suren Baghdasaryan

Replace BUG_ON(vma->vm_flags & VM_PFNMAP) in vm_insert_page with WARN_ON_ONCE and returning an error. This is to ensure users of the vm_insert_page that set VM_PFNMAP are notified of the wrong flag usage and get an indication of an error without panicing the kernel. This will help identifying drivers that need to clear VM_PFNMAP before using dmabuf system heap which is moving to use vm_insert_page. Suggested-by: Christoph Hellwig <hch(a)infradead.org> Signed-off-by: Suren Baghdasaryan <surenb(a)google.com> --- mm/memory.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/mm/memory.c b/mm/memory.c index feff48e1465a..e503c9801cd9 100644 --- a/mm/memory.c +++ b/mm/memory.c @@ -1827,7 +1827,8 @@ int vm_insert_page(struct vm_area_struct *vma, unsigned long addr, return -EINVAL; if (!(vma->vm_flags & VM_MIXEDMAP)) { BUG_ON(mmap_read_trylock(vma->vm_mm)); - BUG_ON(vma->vm_flags & VM_PFNMAP); + if (WARN_ON_ONCE(vma->vm_flags & VM_PFNMAP)) + return -EINVAL; vma->vm_flags |= VM_MIXEDMAP; } return insert_page(vma, addr, page, vma->vm_page_prot); -- 2.30.0.365.g02bc693789-goog

4 years, 9 months

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

Linaro-mm-sig