DRM GPUVA Manager & Nouveau VM_BIND UAPI
Subsystems (4)
v6 cover DRM GPUVA Manager & Nouveau VM_BIND UAPI
This patch series provides a new UAPI for the Nouveau driver in order to support Vulkan features, such as sparse bindings and sparse residency.
Furthermore, with the DRM GPUVA manager it provides a new DRM core feature to keep track of GPU virtual address (VA) mappings in a more generic way.
The DRM GPUVA manager is indented to help drivers implement userspace-manageable GPU VA spaces in reference to the Vulkan API. In order to achieve this goal it serves the following purposes in this context.
1) Provide infrastructure to track GPU VA allocations and mappings, making use of the maple_tree.
2) Generically connect GPU VA mappings to their backing buffers, in particular DRM GEM objects.
3) Provide a common implementation to perform more complex mapping operations on the GPU VA space. In particular splitting and merging of GPU VA mappings, e.g. for intersecting mapping requests or partial unmap requests.
The new VM_BIND Nouveau UAPI build on top of the DRM GPUVA manager, itself providing the following new interfaces.
1) Initialize a GPU VA space via the new DRM_IOCTL_NOUVEAU_VM_INIT ioctl for UMDs to specify the portion of VA space managed by the kernel and userspace, respectively.
2) Allocate and free a VA space region as well as bind and unbind memory to the GPUs VA space via the new DRM_IOCTL_NOUVEAU_VM_BIND ioctl.
3) Execute push buffers with the new DRM_IOCTL_NOUVEAU_EXEC ioctl.
Both, DRM_IOCTL_NOUVEAU_VM_BIND and DRM_IOCTL_NOUVEAU_EXEC, make use of the DRM scheduler to queue jobs and support asynchronous processing with DRM syncobjs as synchronization mechanism.
By default DRM_IOCTL_NOUVEAU_VM_BIND does synchronous processing, DRM_IOCTL_NOUVEAU_EXEC supports asynchronous processing only.
The new VM_BIND UAPI for Nouveau makes also use of drm_exec (execution context for GEM buffers) by Christian König. Since the patch implementing drm_exec was not yet merged into drm-next it is part of this series, as well as a small fix for this patch, which was found while testing this series.
This patch series is also available at [1].
There is a Mesa NVK merge request by Dave Airlie [2] implementing the corresponding userspace parts for this series.
The Vulkan CTS test suite passes the sparse binding and sparse residency test cases for the new UAPI together with Dave's Mesa work.
There are also some test cases in the igt-gpu-tools project [3] for the new UAPI and hence the DRM GPU VA manager. However, most of them are testing the DRM GPU VA manager's logic through Nouveau's new UAPI and should be considered just as helper for implementation.
However, I absolutely intend to change those test cases to proper kunit test cases for the DRM GPUVA manager, once and if we agree on it's usefulness and design.
[1] https://gitlab.freedesktop.org/nouvelles/kernel/-/tree/new-uapi-drm-next /
https://gitlab.freedesktop.org/nouvelles/kernel/-/merge_requests/1
[2] https://gitlab.freedesktop.org/nouveau/mesa/-/merge_requests/150/
[3] https://gitlab.freedesktop.org/dakr/igt-gpu-tools/-/tree/wip_nouveau_vm_bind
Changes in V2:
==============
Nouveau:
- Reworked the Nouveau VM_BIND UAPI to avoid memory allocations in fence signalling critical sections. Updates to the VA space are split up in three separate stages, where only the 2. stage executes in a fence signalling critical section:
1. update the VA space, allocate new structures and page tables
2. (un-)map the requested memory bindings
3. free structures and page tables
- Separated generic job scheduler code from specific job implementations.
- Separated the EXEC and VM_BIND implementation of the UAPI.
- Reworked the locking parts of the nvkm/vmm RAW interface, such that (un-)map operations can be executed in fence signalling critical sections.
GPUVA Manager:
- made drm_gpuva_regions optional for users of the GPUVA manager
- allow NULL GEMs for drm_gpuva entries
- swichted from drm_mm to maple_tree for track drm_gpuva / drm_gpuva_region entries
- provide callbacks for users to allocate custom drm_gpuva_op structures to allow inheritance
- added user bits to drm_gpuva_flags
- added a prefetch operation type in order to support generating prefetch operations in the same way other operations generated
- hand the responsibility for mutual exclusion for a GEM's drm_gpuva list to the user; simplified corresponding (un-)link functions
Maple Tree:
- I added two maple tree patches to the series, one to support custom tree walk macros and one to hand the locking responsibility to the user of the GPUVA manager without pre-defined lockdep checks.
Changes in V3:
==============
Nouveau:
- Reworked the Nouveau VM_BIND UAPI to do the job cleanup (including page table cleanup) within a workqueue rather than the job_free() callback of the scheduler itself. A job_free() callback can stall the execution (run() callback) of the next job in the queue. Since the page table cleanup requires to take the same locks as need to be taken for page table allocation, doing it directly in the job_free() callback would still violate the fence signalling critical path.
- Separated Nouveau fence allocation and emit, such that we do not violate the fence signalling critical path in EXEC jobs.
- Implement "regions" (for handling sparse mappings through PDEs and dual page tables) within Nouveau.
- Drop the requirement for every mapping to be contained within a region.
- Add necassary synchronization of VM_BIND job operation sequences in order to work around limitations in page table handling. This will be addressed in a future re-work of Nouveau's page table handling.
- Fixed a couple of race conditions found through more testing. Thanks to Dave for consitently trying to break it. :-)
GPUVA Manager:
- Implement pre-allocation capabilities for tree modifications within fence signalling critical sections.
- Implement accessors to to apply tree modification while walking the GPUVA tree in order to actually support processing of drm_gpuva_ops through callbacks in fence signalling critical sections rather than through pre-allocated operation lists.
- Remove merging of GPUVAs; the kernel has limited to none knowlege about the semantics of mapping sequences. Hence, merging is purely speculative. It seems that gaining a significant (or at least a measurable) performance increase through merging is way more likely to happen when userspace is responsible for merging mappings up to the next larger page size if possible.
- Since merging was removed, regions pretty much loose their right to exist. They might still be useful for handling dual page tables or similar mechanisms, but since Nouveau seems to be the only driver having a need for this for now, regions were removed from the GPUVA manager.
- Fixed a couple of maple_tree related issues; thanks to Liam for helping me out.
Changes in V4:
==============
Nouveau:
- Refactored how specific VM_BIND and EXEC jobs are created and how their arguments are passed to the generic job implementation.
- Fixed a UAF race condition where bind job ops could have been freed already while still waiting for a job cleanup to finish. This is due to in certain cases we need to wait for mappings actually being unmapped before creating sparse regions in the same area.
- Re-based the code onto drm_exec v4 patch.
GPUVA Manager:
- Fixed a maple tree related bug when pre-allocating MA states. (Boris Brezillion)
- Made struct drm_gpuva_fn_ops a const object in all occurrences. (Boris Brezillion)
Changes in V5:
==============
Nouveau:
- Link and unlink GPUVAs outside the fence signalling critical path in nouveau_uvmm_bind_job_submit() holding the dma-resv lock. Mutual exclusion of BO evicts causing mapping invalidation and regular mapping operations is ensured with dma-fences.
GPUVA Manager:
- Removed the separate GEMs GPUVA list lock. Link and unlink as well as iterating the GEM's GPUVA list should be protected with the GEM's dma-resv lock instead.
- Renamed DRM_GPUVA_EVICTED flag to DRM_GPUVA_INVALIDATED. Mappings do not get eviced, they might get invalidated due to eviction.
- Maple tree uses the 'unsinged long' type for node entries. While this works for GPU VA spaces larger than 32-bit on 64-bit kernel, the GPU VA space is limited to 32-bit on 32-bit kernels as well.
As long as we do not have a 64-bit capable maple tree for 32-bit kernels, the GPU VA manager contains checks to throw warnings when GPU VA entries exceed the maple tree's storage capabilities.
- Extended the Documentation and added example code as requested by Donald Robson.
Changes in V6
=============
Nouveau:
- Re-based the code onto drm_exec v5 patch.
GPUVA Manager:
- Switch from maple tree to RB-tree.
It turned out that mas_preallocate() requires the maple tree not to change in between pre-allocating nodes with mas_preallocate() and inserting an entry with the help of the pre-allocated memory (mas_insert_prealloc()).
However, considering that drivers typically implement interfaces where jobs to create GPU mappings can be submitted by userspace, are queued up by the kernel and are processed asynchronously in dma-fence signalling critical paths, this is a major issue. In the ioctl() used to submit a job we'd need to pre-allocated memory with mas_preallocate(), however, previously queued up jobs could concurrently alter the maple tree resulting in potentially insufficient pre-allocated memory for the currently submitted job on execution time.
There is a detailed and still ongoing discussion about this topic one the -mm list [1]. So far the only solution seems to be to use GFP_ATOMIC and allocate memory directly in the fence signalling critical path, where we need it. However, I think that is not what we want to rely on.
I think we should definitely continue in trying to find a solution on how to fit in the maple tree (or how to make the maple tree fit in). However, for now it seems to be more expedient to move on using a RB-tree.
[1] https://lore.kernel.org/lkml/[email protected]/
- Provide a flag to let driver optionally provide their own lock to lock linking and unlinking of GPUVAs to GEM objects. The DRM GPUVA manager still does not take the locks itself, but rather contains lockdep checks on either the GEMs dma-resv lock (default) or, if DRM_GPUVA_MANAGER_LOCK_EXTERN is set, the driver provided lock. (Boris Brezillon)
Christian König (1):
drm: execution context for GEM buffers v5
Danilo Krummrich (12):
drm: manager to keep track of GPUs VA mappings
drm: debugfs: provide infrastructure to dump a DRM GPU VA space drm/nouveau: new VM_BIND uapi interfaces
drm/nouveau: get vmm via nouveau_cli_vmm() drm/nouveau: bo: initialize GEM GPU VA interface drm/nouveau: move usercopy helpers to nouveau_drv.h drm/nouveau: fence: separate fence alloc and emit drm/nouveau: fence: fail to emit when fence context is killed drm/nouveau: chan: provide nouveau_channel_kill() drm/nouveau: nvkm/vmm: implement raw ops to manage uvmm drm/nouveau: implement new VM_BIND uAPI
drm/nouveau: debugfs: implement DRM GPU VA debugfs
Documentation/gpu/driver-uapi.rst | 11 + Documentation/gpu/drm-mm.rst | 48 + drivers/gpu/drm/Kconfig | 6 + drivers/gpu/drm/Makefile | 3 + drivers/gpu/drm/drm_debugfs.c | 40 + drivers/gpu/drm/drm_exec.c | 330 +++ drivers/gpu/drm/drm_gem.c | 3 + drivers/gpu/drm/drm_gpuva_mgr.c | 1743 +++++++++++++++ drivers/gpu/drm/nouveau/Kbuild | 3 + drivers/gpu/drm/nouveau/Kconfig | 2 + drivers/gpu/drm/nouveau/dispnv04/crtc.c | 9 +- drivers/gpu/drm/nouveau/include/nvif/if000c.h | 26 +- drivers/gpu/drm/nouveau/include/nvif/vmm.h | 19 +- .../gpu/drm/nouveau/include/nvkm/subdev/mmu.h | 20 +- drivers/gpu/drm/nouveau/nouveau_abi16.c | 24 + drivers/gpu/drm/nouveau/nouveau_abi16.h | 1 + drivers/gpu/drm/nouveau/nouveau_bo.c | 204 +- drivers/gpu/drm/nouveau/nouveau_bo.h | 2 +- drivers/gpu/drm/nouveau/nouveau_chan.c | 22 +- drivers/gpu/drm/nouveau/nouveau_chan.h | 1 + drivers/gpu/drm/nouveau/nouveau_debugfs.c | 39 + drivers/gpu/drm/nouveau/nouveau_dmem.c | 9 +- drivers/gpu/drm/nouveau/nouveau_drm.c | 27 +- drivers/gpu/drm/nouveau/nouveau_drv.h | 94 +- drivers/gpu/drm/nouveau/nouveau_exec.c | 414 ++++ drivers/gpu/drm/nouveau/nouveau_exec.h | 54 + drivers/gpu/drm/nouveau/nouveau_fence.c | 23 +- drivers/gpu/drm/nouveau/nouveau_fence.h | 5 +- drivers/gpu/drm/nouveau/nouveau_gem.c | 62 +- drivers/gpu/drm/nouveau/nouveau_mem.h | 5 + drivers/gpu/drm/nouveau/nouveau_prime.c | 2 +- drivers/gpu/drm/nouveau/nouveau_sched.c | 462 ++++ drivers/gpu/drm/nouveau/nouveau_sched.h | 123 + drivers/gpu/drm/nouveau/nouveau_svm.c | 2 +- drivers/gpu/drm/nouveau/nouveau_uvmm.c | 1970 +++++++++++++++++ drivers/gpu/drm/nouveau/nouveau_uvmm.h | 107 + drivers/gpu/drm/nouveau/nouveau_vmm.c | 4 +- drivers/gpu/drm/nouveau/nvif/vmm.c | 100 +- .../gpu/drm/nouveau/nvkm/subdev/mmu/uvmm.c | 213 +- drivers/gpu/drm/nouveau/nvkm/subdev/mmu/vmm.c | 197 +- drivers/gpu/drm/nouveau/nvkm/subdev/mmu/vmm.h | 25 + .../drm/nouveau/nvkm/subdev/mmu/vmmgf100.c | 16 +- .../drm/nouveau/nvkm/subdev/mmu/vmmgp100.c | 16 +- .../gpu/drm/nouveau/nvkm/subdev/mmu/vmmnv50.c | 27 +- include/drm/drm_debugfs.h | 25 + include/drm/drm_drv.h | 6 + include/drm/drm_exec.h | 120 + include/drm/drm_gem.h | 52 + include/drm/drm_gpuva_mgr.h | 756 +++++++ include/uapi/drm/nouveau_drm.h | 209 ++ 50 files changed, 7441 insertions(+), 240 deletions(-) create mode 100644 drivers/gpu/drm/drm_exec.c create mode 100644 drivers/gpu/drm/drm_gpuva_mgr.c create mode 100644 drivers/gpu/drm/nouveau/nouveau_exec.c create mode 100644 drivers/gpu/drm/nouveau/nouveau_exec.h create mode 100644 drivers/gpu/drm/nouveau/nouveau_sched.c create mode 100644 drivers/gpu/drm/nouveau/nouveau_sched.h create mode 100644 drivers/gpu/drm/nouveau/nouveau_uvmm.c create mode 100644 drivers/gpu/drm/nouveau/nouveau_uvmm.h create mode 100644 include/drm/drm_exec.h create mode 100644 include/drm/drm_gpuva_mgr.h
base-commit: 5ff2977b19769fd24b0cfbe7cbe4d5114b6106af
Signature · 1 line
2.41.0
PATCH v6 1/13 drm: execution context for GEM buffers v5
From: Christian König <[email protected]>
This adds the infrastructure for an execution context for GEM buffers which is similar to the existing TTMs execbuf util and intended to replace it in the long term.
The basic functionality is that we abstracts the necessary loop to lock many different GEM buffers with automated deadlock and duplicate handling.
v2: drop xarray and use dynamic resized array instead, the locking
overhead is unecessary and measurable.
v3: drop duplicate tracking, radeon is really the only one needing that.
v4: fixes issues pointed out by Danilo, some typos in comments and a
helper for lock arrays of GEM objects.
v5: some suggestions by Boris Brezillon, especially just use one retry
macro, drop loop in prepare_array, use flags instead of bool
--- Documentation/gpu/drm-mm.rst | 12 ++ drivers/gpu/drm/Kconfig | 6 + drivers/gpu/drm/Makefile | 2 + drivers/gpu/drm/drm_exec.c | 330 +++++++++++++++++++++++++++++++++++ include/drm/drm_exec.h | 120 +++++++++++++ 5 files changed, 470 insertions(+) create mode 100644 drivers/gpu/drm/drm_exec.c create mode 100644 include/drm/drm_exec.h
Documentation/gpu/drm-mm.rst +12 -0
@@ -493,6 +493,18 @@ DRM Sync Objects .. kernel-doc:: drivers/gpu/drm/drm_syncobj.c :export: +DRM Execution context +===================== + +.. kernel-doc:: drivers/gpu/drm/drm_exec.c + :doc: Overview + +.. kernel-doc:: include/drm/drm_exec.h + :internal: + +.. kernel-doc:: drivers/gpu/drm/drm_exec.c + :export: + GPU Scheduler =============
drivers/gpu/drm/Kconfig +6 -0
@@ -194,6 +194,12 @@ config DRM_TTM GPU memory types. Will be enabled automatically if a device driver uses it. +config DRM_EXEC + tristate + depends on DRM + help + Execution context for command submissions + config DRM_BUDDY tristate depends on DRM
drivers/gpu/drm/Makefile +2 -0
@@ -78,6 +78,8 @@ obj-$(CONFIG_DRM_PANEL_ORIENTATION_QUIRKS) += drm_panel_orientation_quirks.o # # Memory-management helpers # +# +obj-$(CONFIG_DRM_EXEC) += drm_exec.o obj-$(CONFIG_DRM_BUDDY) += drm_buddy.o
drivers/gpu/drm/drm_exec.c new file +330 -0
330 lines not on this page · raw message
include/drm/drm_exec.h new file +120 -0
120 lines not on this page · raw message
Signature · 1 line
2.41.0
PATCH v6 2/13 drm: manager to keep track of GPUs VA mappings
In reply to Danilo KrummrichAdd infrastructure to keep track of GPU virtual address (VA) mappings with a decicated VA space manager implementation.
New UAPIs, motivated by Vulkan sparse memory bindings graphics drivers start implementing, allow userspace applications to request multiple and arbitrary GPU VA mappings of buffer objects. The DRM GPU VA manager is intended to serve the following purposes in this context.
1) Provide infrastructure to track GPU VA allocations and mappings, making use of the maple_tree.
2) Generically connect GPU VA mappings to their backing buffers, in particular DRM GEM objects.
3) Provide a common implementation to perform more complex mapping operations on the GPU VA space. In particular splitting and merging of GPU VA mappings, e.g. for intersecting mapping requests or partial unmap requests.
--- Documentation/gpu/drm-mm.rst | 36 + drivers/gpu/drm/Makefile | 1 + drivers/gpu/drm/drm_gem.c | 3 + drivers/gpu/drm/drm_gpuva_mgr.c | 1743 +++++++++++++++++++++++++++++++ include/drm/drm_drv.h | 6 + include/drm/drm_gem.h | 52 + include/drm/drm_gpuva_mgr.h | 756 ++++++++++++++ 7 files changed, 2597 insertions(+) create mode 100644 drivers/gpu/drm/drm_gpuva_mgr.c create mode 100644 include/drm/drm_gpuva_mgr.h
Documentation/gpu/drm-mm.rst +36 -0
@@ -466,6 +466,42 @@ DRM MM Range Allocator Function References .. kernel-doc:: drivers/gpu/drm/drm_mm.c :export: +DRM GPU VA Manager +================== + +Overview +-------- + +.. kernel-doc:: drivers/gpu/drm/drm_gpuva_mgr.c + :doc: Overview + +Split and Merge +--------------- + +.. kernel-doc:: drivers/gpu/drm/drm_gpuva_mgr.c + :doc: Split and Merge + +Locking +------- + +.. kernel-doc:: drivers/gpu/drm/drm_gpuva_mgr.c + :doc: Locking + +Examples +-------- + +.. kernel-doc:: drivers/gpu/drm/drm_gpuva_mgr.c + :doc: Examples + +DRM GPU VA Manager Function References +-------------------------------------- + +.. kernel-doc:: include/drm/drm_gpuva_mgr.h + :internal: + +.. kernel-doc:: drivers/gpu/drm/drm_gpuva_mgr.c + :export: + DRM Buddy Allocator ===================
drivers/gpu/drm/Makefile +1 -0
@@ -45,6 +45,7 @@ drm-y := \ drm_vblank.o \ drm_vblank_work.o \ drm_vma_manager.o \ + drm_gpuva_mgr.o \ drm_writeback.o drm-$(CONFIG_DRM_LEGACY) += \ drm_agpsupport.o \
drivers/gpu/drm/drm_gem.c +3 -0
@@ -164,6 +164,9 @@ void drm_gem_private_object_init(struct drm_device *dev,if (!obj->resv) obj->resv = &obj->_resv; +if (drm_core_check_feature (dev,DRIVER_GEM_GPUVA )) +drm_gem_gpuva_init (obj); +drm_vma_node_reset (&obj->vma_node);INIT_LIST_HEAD (&obj->lru_node); }
drivers/gpu/drm/drm_gpuva_mgr.c new file +1743 -0
1,743 lines not on this page · raw message
include/drm/drm_drv.h +6 -0
@@ -104,6 +104,12 @@ enum drm_driver_feature {* acceleration should be handled by two drivers that are connected using auxiliary bus. */ DRIVER_COMPUTE_ACCEL =BIT (7 ), +/** +* @DRIVER_GEM_GPUVA: +* +* Driver supports user defined GPU VA bindings for GEM objects. +*/ + DRIVER_GEM_GPUVA = BIT(8),/* IMPORTANT: Below are all the legacy flags, add new ones above. */
include/drm/drm_gem.h +52 -0
68 lines not on this page · raw message
include/drm/drm_gpuva_mgr.h new file +756 -0
756 lines not on this page · raw message
Signature · 1 line
2.41.0
Re: [PATCH drm-next v6 02/13] drm: manager to keep track of GPUs VA mappings
Hi Danilo,
Quoted 11 lines · On Fri, 30 Jun 2023 00:25:18 +0200 Danilo Krummrich <[email protected]> wrote:
+ * int
driver_gpuva_remap(struct drm_gpuva_op *op, void *__ctx)
+ * {
+ * structdriver_context*ctx =__ctx;
+ *
+ *drm_gpuva_remap(ctx->prev_va, ctx->next_va, &op->remap);
+ *
+ *drm_gpuva_unlink(op->remap.unmap->va);
+ *kfree(op->remap.unmap->va);
+ *
+ * if (op->remap.prev) {
+ *drm_gpuva_link(ctx->prev_va);
I ended up switching to dma_resv-based locking for the GEMs and I wonder what the locking is supposed to look like in the async-mapping case, where we insert/remove the VA nodes in the drm_sched::run_job() path.
What I have right now is something like:
dma_resv_lock(vm->resv);
// split done in drm_gpuva_sm_map(), each iteration
// of the loop is a call to the driver ->[re,un]map()
// hook
for_each_sub_op() {
// Private BOs have their resv field pointing to the
// VM resv and we take the VM resv lock before calling
// drm_gpuva_sm_map()
if (vm->resv != gem->resv)
dma_resv_lock(gem->resv);
drm_gpuva_[un]link(va);
gem_[un]pin(gem);
if (vm->resv != gem->resv)
dma_resv_unlock(gem->resv);
}
dma_resv_unlock(vm->resv);In practice, I don't expect things to deadlock, because the VM resv is not supposed to be taken outside the VM context and the locking order is always the same (VM lock first, and then each shared BO taken/released independently), but I'm not super thrilled by this nested lock, and I'm wondering if we shouldn't have a pass collecting locks in a drm_exec context first, and then have
the operations executed. IOW, something like that:
drm_exec_init(exec, DRM_EXEC_IGNORE_DUPLICATES)
drm_exec_until_all_locked(exec) {
// Dummy GEM is the dummy GEM object I use to make the VM
// participate in the locking without having to teach
// drm_exec how to deal with raw dma_resv objects.
ret = drm_exec_lock_obj(exec, vm->dummy_gem);
drm_exec_retry_on_contention(exec);
if (ret)
return ret;
// Could take the form of drm_gpuva_sm_[un]map_acquire_locks()
// helpers
for_each_sub_op() {
ret = drm_exec_lock_obj(exec, gem);
if (ret)
return ret;
}
}
// each iteration of the loop is a call to the driver
// ->[re,un]map() hook
for_each_sub_op() {
...
gem_[un]pin_locked(gem);
drm_gpuva_[un]link(va);
...
}
drm_exec_fini(exec);Don't know if I got this right, or if I'm just confused again by how the drm_gpuva API is supposed to be used.
Regards,
Boris
Quoted 10 lines
+ *
ctx->prev_va= NULL;
+ * }
+ *
+ * if (op->remap.next) {
+ *drm_gpuva_link(ctx->next_va);
+ *ctx->next_va= NULL;
+ * }
+ *
+ * return 0;
+ * }
Re: [PATCH drm-next v6 02/13] drm: manager to keep track of GPUs VA mappings
Quoted 70 lines · On Fri, 30 Jun 2023 10:02:52 +0200 Boris Brezillon…
Hi Danilo,
Quoted 11 lines · On Fri, 30 Jun 2023 00:25:18 +0200 Danilo Krummrich <[email protected]> wrote:
+ * int
driver_gpuva_remap(struct drm_gpuva_op *op, void *__ctx)
+ * {
+ * structdriver_context*ctx =__ctx;
+ *
+ *drm_gpuva_remap(ctx->prev_va, ctx->next_va, &op->remap);
+ *
+ *drm_gpuva_unlink(op->remap.unmap->va);
+ *kfree(op->remap.unmap->va);
+ *
+ * if (op->remap.prev) {
+ *drm_gpuva_link(ctx->prev_va);I ended up switching to dma_resv-based locking for the GEMs and I wonder what the locking is supposed to look like in the async-mapping case, where we insert/remove the VA nodes in the
drm_sched::run_job()path.What I have right now is something like:
dma_resv_lock(vm->resv); // split done in drm_gpuva_sm_map(), each iteration // of the loop is a call to the driver ->[re,un]map() // hook for_each_sub_op() { // Private BOs have their resv field pointing to the // VM resv and we take the VM resv lock before calling // drm_gpuva_sm_map() if (vm->resv != gem->resv) dma_resv_lock(gem->resv); drm_gpuva_[un]link(va); gem_[un]pin(gem); if (vm->resv != gem->resv) dma_resv_unlock(gem->resv); } dma_resv_unlock(vm->resv);In practice, I don't expect things to deadlock, because the VM resv is not supposed to be taken outside the VM context and the locking order is always the same (VM lock first, and then each shared BO taken/released independently), but I'm not super thrilled by this nested lock, and I'm wondering if we shouldn't have a pass collecting locks in a
drm_execcontext first, and then have
the operations executed. IOW, something like that:drm_exec_init(exec, DRM_EXEC_IGNORE_DUPLICATES) drm_exec_until_all_locked(exec) { // Dummy GEM is the dummy GEM object I use to make the VM // participate in the locking without having to teach // drm_exec how to deal with raw dma_resv objects. ret = drm_exec_lock_obj(exec, vm->dummy_gem); drm_exec_retry_on_contention(exec); if (ret) return ret; // Could take the form of drm_gpuva_sm_[un]map_acquire_locks() // helpers for_each_sub_op() { ret = drm_exec_lock_obj(exec, gem); if (ret) return ret;
Quoted 8 lines
}
}
// each iteration of the loop is a call to the driver
// ->[re,un]map() hook
for_each_sub_op() {
...
gem_[un]pin_locked(gem);Just wanted to clarify that the pages have been pinned at VM_BIND job creation time, so this gem_pin_locked() call is effectively just a pin_count++, not the whole page allocation, which we don't want to happen in a dma-signaling path.
Quoted 5 lines
drm_gpuva_[un]link(va); ... } drm_exec_fini(exec);
Re: [PATCH drm-next v6 02/13] drm: manager to keep track of GPUs VA mappings
In reply to Boris BrezillonQuoted 8 lines · On Fri, 30 Jun 2023 10:02:52 +0200 Boris Brezillon…
In practice, I don't expect things to deadlock, because the VM resv is not supposed to be taken outside the VM context and the locking order is always the same (VM lock first, and then each shared BO taken/released independently), but I'm not super thrilled by this nested lock, and I'm wondering if we shouldn't have a pass collecting locks in a
drm_execcontext first, and then have
the operations executed. IOW, something like that:
Quoted 12 lines
drm_exec_init(exec, DRM_EXEC_IGNORE_DUPLICATES)
drm_exec_until_all_locked(exec) {
// Dummy GEM is the dummy GEM object I use to make the VM
// participate in the locking without having to teach
// drm_exec how to deal with raw dma_resv objects.
ret = drm_exec_lock_obj(exec, vm->dummy_gem);
drm_exec_retry_on_contention(exec);
if (ret)
return ret;
// Could take the form of drm_gpuva_sm_[un]map_acquire_locks()
// helpersNevermind, I implemented a driver specific acquire_op_locks(), and it's fairly simple with the gpuva iter (we just have to iterate over all VAs covered by the operation range and call drm_exec_lock_obj() on the GEM attached to these VAs), so it's probably not worth providing a generic helper for that.
Re: [PATCH drm-next v6 02/13] drm: manager to keep track of GPUs VA mappings
In reply to Boris BrezillonHi Boris,
Quoted 17 lines · Boris Brezillon
Hi Danilo,
Quoted 11 lines · On Fri, 30 Jun 2023 00:25:18 +0200 Danilo Krummrich <[email protected]> wrote:
+ * int
driver_gpuva_remap(struct drm_gpuva_op *op, void *__ctx)
+ * {
+ * structdriver_context*ctx =__ctx;
+ *
+ *drm_gpuva_remap(ctx->prev_va, ctx->next_va, &op->remap);
+ *
+ *drm_gpuva_unlink(op->remap.unmap->va);
+ *kfree(op->remap.unmap->va);
+ *
+ * if (op->remap.prev) {
+ *drm_gpuva_link(ctx->prev_va);
Quoted 4 lines
I ended up switching to dma_resv-based locking for the GEMs and I wonder what the locking is supposed to look like in the async-mapping case, where we insert/remove the VA nodes in the
drm_sched::run_job()path.
If you decide to pick the interface where you just call drm_gpuva_sm_[un]map() and receive a callback for each operation it takes to fulfill the request, you probably do this because you want to do everything one shot, updating the VA space, link/unlink GPUVAs to/from its corresponding backing GEMs, do the actual GPU mappings.
This has a few advantages over generating a list of operations when the job is submitted. You've pointed out one of them, when you noticed that with a list of operations one can't sneak in a synchronous job between already queued up asynchronous jobs.
However, for the asynchronous path it has the limitation that the dma-resv lock can't be used to link/unlink GPUVAs to/from its corresponding backing GEMs, since this would happen in the fence signalling critical path and we're not allowed to hold the dma-resv lock there. Hence, as we discussed I added the option for drivers to provide an external lock for that, just to be able to keep some lockdep checks.
Quoted 16 lines
What I have right now is something like:
dma_resv_lock(vm->resv); // split done in drm_gpuva_sm_map(), each iteration // of the loop is a call to the driver ->[re,un]map() // hook for_each_sub_op() { // Private BOs have their resv field pointing to the // VM resv and we take the VM resv lock before calling // drm_gpuva_sm_map() if (vm->resv != gem->resv) dma_resv_lock(gem->resv);
Quoted 9 lines
drm_gpuva_[un]link(va); gem_[un]pin(gem); if (vm->resv != gem->resv) dma_resv_unlock(gem->resv); } dma_resv_unlock(vm->resv);
I'm not sure I get this code right, reading "for_each_sub_op()" and "drm_gpuva_sm_map()" looks a bit like things are mixed up?
Or do you mean to represent the sum of all callbacks with "for_each_sub_op()"? In this case I assume this code runs in drm_sched::run_job() and hence isn't allowed to take the dma-resv lock.
Quoted 29 lines
In practice, I don't expect things to deadlock, because the VM resv is not supposed to be taken outside the VM context and the locking order is always the same (VM lock first, and then each shared BO taken/released independently), but I'm not super thrilled by this nested lock, and I'm wondering if we shouldn't have a pass collecting locks in a
drm_execcontext first, and then have
the operations executed. IOW, something like that:drm_exec_init(exec, DRM_EXEC_IGNORE_DUPLICATES) drm_exec_until_all_locked(exec) { // Dummy GEM is the dummy GEM object I use to make the VM // participate in the locking without having to teach // drm_exec how to deal with raw dma_resv objects. ret = drm_exec_lock_obj(exec, vm->dummy_gem); drm_exec_retry_on_contention(exec); if (ret) return ret; // Could take the form of drm_gpuva_sm_[un]map_acquire_locks() // helpers for_each_sub_op() { ret = drm_exec_lock_obj(exec, gem); if (ret) return ret; } } // each iteration of the loop is a call to the driver // ->[re,un]map() hook
Quoted 8 lines
for_each_sub_op() {
...
gem_[un]pin_locked(gem);
drm_gpuva_[un]link(va);
...
}
drm_exec_fini(exec);I have a follow-up patch (still WIP) in the queue to generalize dma-resv handling, fence handling and GEM validation within the GPUVA manager as optional helper functions: https://gitlab.freedesktop.org/nouvelles/kernel/-/commit/a5fc29f3b1edbf3f96fb5a21b858ffe00a3f2584
This was suggested by Matt Brost.
- Danilo
Quoted 19 lines
Don't know if I got this right, or if I'm just confused again by how the
drm_gpuvaAPI is supposed to be used.Regards,
Boris
Quoted 10 lines
+ *
ctx->prev_va= NULL;
+ * }
+ *
+ * if (op->remap.next) {
+ *drm_gpuva_link(ctx->next_va);
+ *ctx->next_va= NULL;
+ * }
+ *
+ * return 0;
+ * }
Re: [PATCH drm-next v6 02/13] drm: manager to keep track of GPUs VA mappings
Quoted 36 lines · On Thu, 6 Jul 2023 17:06:08 +0200 Danilo Krummrich <[email protected]> wrote:
Hi Boris,
Quoted 17 lines · Boris Brezillon
Hi Danilo,
Quoted 11 lines · On Fri, 30 Jun 2023 00:25:18 +0200 Danilo Krummrich <[email protected]> wrote:
+ * int
driver_gpuva_remap(struct drm_gpuva_op *op, void *__ctx)
+ * {
+ * structdriver_context*ctx =__ctx;
+ *
+ *drm_gpuva_remap(ctx->prev_va, ctx->next_va, &op->remap);
+ *
+ *drm_gpuva_unlink(op->remap.unmap->va);
+ *kfree(op->remap.unmap->va);
+ *
+ * if (op->remap.prev) {
+ *drm_gpuva_link(ctx->prev_va);Quoted 4 lines
I ended up switching to dma_resv-based locking for the GEMs and I wonder what the locking is supposed to look like in the async-mapping case, where we insert/remove the VA nodes in the
drm_sched::run_job()path.If you decide to pick the interface where you just call drm_gpuva_sm_[un]map() and receive a callback for each operation it takes to fulfill the request, you probably do this because you want to do everything one shot, updating the VA space, link/unlink GPUVAs to/from its corresponding backing GEMs, do the actual GPU mappings.
This has a few advantages over generating a list of operations when the job is submitted. You've pointed out one of them, when you noticed that with a list of operations one can't sneak in a synchronous job between already queued up asynchronous jobs.
Quoted 6 lines
However, for the asynchronous path it has the limitation that the dma-resv lock can't be used to link/unlink GPUVAs to/from its corresponding backing GEMs, since this would happen in the fence signalling critical path and we're not allowed to hold the dma-resv lock there. Hence, as we discussed I added the option for drivers to provide an external lock for that, just to be able to keep some lockdep checks.
Uh, okay, I guess that means I need to go back to a custom lock for VM operations then.
Quoted 24 lines
Quoted 24 lines
What I have right now is something like:
dma_resv_lock(vm->resv); // split done in drm_gpuva_sm_map(), each iteration // of the loop is a call to the driver ->[re,un]map() // hook for_each_sub_op() { // Private BOs have their resv field pointing to the // VM resv and we take the VM resv lock before calling // drm_gpuva_sm_map() if (vm->resv != gem->resv) dma_resv_lock(gem->resv); drm_gpuva_[un]link(va); gem_[un]pin(gem); if (vm->resv != gem->resv) dma_resv_unlock(gem->resv); }
Quoted 8 lines
Quoted 2 lines
dma_resv_unlock(vm->resv);I'm not sure I get this code right, reading "
for_each_sub_op()" and "drm_gpuva_sm_map()" looks a bit like things are mixed up?Or do you mean to represent the sum of all callbacks with "
for_each_sub_op()"?
That ^.
Quoted 2 lines
In this case I assume this code runs in
drm_sched::run_job()and hence isn't allowed to take the dma-resv lock.
Yeah, I didn't realize that taking the dma-resv lock in the dma-signaling path was forbidden. I think it's fine for the drm_gpuva destroy operation (which calls drm_gem_shmem_unpin(), which in turns acquires the resv lock) because I can move that to a worker and get it out of the dma-signaling path. The problem remains for remap operations though. I need to call drm_gem_shmem_pin() so we retain the pages even after the unmapped gpuva object that's in the middle of a mapping is released. I guess one option would be to use an atomic_t for drm_shmem_gem_object::pages_use_count, and
have something like:
int drm_gem_shmem_pin(struct drm_gem_shmem_object *shmem)
{
int ret;
if (atomic_inc_not_zero(&shmem->pages_use_count))
return 0;
dma_resv_lock(shmem->base.resv, NULL);
ret = drm_gem_shmem_pin_locked(shmem);
dma_resv_unlock(shmem->base.resv);
return ret;
}Given the object already had its pages pinned when we remap, we're sure the fast path will be taken, and no dma-resv lock aquired.
Quoted 37 lines
Quoted 37 lines
In practice, I don't expect things to deadlock, because the VM resv is not supposed to be taken outside the VM context and the locking order is always the same (VM lock first, and then each shared BO taken/released independently), but I'm not super thrilled by this nested lock, and I'm wondering if we shouldn't have a pass collecting locks in a
drm_execcontext first, and then have
the operations executed. IOW, something like that:drm_exec_init(exec, DRM_EXEC_IGNORE_DUPLICATES) drm_exec_until_all_locked(exec) { // Dummy GEM is the dummy GEM object I use to make the VM // participate in the locking without having to teach // drm_exec how to deal with raw dma_resv objects. ret = drm_exec_lock_obj(exec, vm->dummy_gem); drm_exec_retry_on_contention(exec); if (ret) return ret; // Could take the form of drm_gpuva_sm_[un]map_acquire_locks() // helpers for_each_sub_op() { ret = drm_exec_lock_obj(exec, gem); if (ret) return ret; } } // each iteration of the loop is a call to the driver // ->[re,un]map() hook for_each_sub_op() { ... gem_[un]pin_locked(gem); drm_gpuva_[un]link(va); ... }
Quoted 6 lines
Quoted 1 line
drm_exec_fini(exec);I have a follow-up patch (still WIP) in the queue to generalize dma-resv handling, fence handling and GEM validation within the GPUVA manager as optional helper functions:
https://gitlab.freedesktop.org/nouvelles/kernel/-/commit/a5fc29f3b1edbf3f96fb5a21b858ffe00a3f2584
Thanks for the heads-up. That's more or less what I have, except I'm attaching a dummy_gem object to the VM so it can be passed to drm_exec directly (instead of having a separate ww_acquire_ctx).
Re: [PATCH drm-next v6 02/13] drm: manager to keep track of GPUs VA mappings
In reply to Danilo KrummrichHi, Danilo
Some review comments below:
Quoted 10 lines · Danilo Krummrich
Add infrastructure to keep track of GPU virtual address (VA) mappings with a decicated VA space manager implementation.
New UAPIs, motivated by Vulkan sparse memory bindings graphics drivers start implementing, allow userspace applications to request multiple and arbitrary GPU VA mappings of buffer objects. The DRM GPU VA manager is intended to serve the following purposes in this context.
1) Provide infrastructure to track GPU VA allocations and mappings, making use of the
maple_tree.
It looks like we're not using the maple tree anymore, but rather an instantiation of an interval tree.
(Perhaps as a follow-up it makes sense to provide a pre-instantiated common u64 version of the interval tree in addition to the unsigned long one since it appears to be used in multiple places in graphics drivers).
Quoted 103 lines
2) Generically connect GPU VA mappings to their backing buffers, in particular DRM GEM objects.
3) Provide a common implementation to perform more complex mapping operations on the GPU VA space. In particular splitting and merging of GPU VA mappings, e.g. for intersecting mapping requests or partial unmap requests.
Tested-by Donald Robson<[email protected]>Reviewed-by Boris Brezillon<[email protected]>Suggested-by Dave Airlie<[email protected]>Signed-off-by Danilo Krummrich<[email protected]>--- Documentation/gpu/drm-mm.rst | 36 + drivers/gpu/drm/Makefile | 1 + drivers/gpu/drm/drm_gem.c | 3 + drivers/gpu/drm/drm_gpuva_mgr.c | 1743 +++++++++++++++++++++++++++++++ include/drm/drm_drv.h | 6 + include/drm/drm_gem.h | 52 + include/drm/drm_gpuva_mgr.h | 756 ++++++++++++++ 7 files changed, 2597 insertions(+)create mode 100644
drivers/gpu/drm/drm_gpuva_mgr.ccreate mode 100644include/drm/drm_gpuva_mgr.hDocumentation/gpu/drm-mm.rst +36
-0@@ -466,6 +466,42 @@ DRM MM Range Allocator Function References .. kernel-doc:: drivers/gpu/drm/drm_mm.c :export: +DRM GPU VA Manager +================== + +Overview +-------- + +.. kernel-doc:: drivers/gpu/drm/drm_gpuva_mgr.c + :doc: Overview + +Split and Merge +--------------- + +.. kernel-doc:: drivers/gpu/drm/drm_gpuva_mgr.c + :doc: Split and Merge + +Locking +------- + +.. kernel-doc:: drivers/gpu/drm/drm_gpuva_mgr.c + :doc: Locking + +Examples +-------- + +.. kernel-doc:: drivers/gpu/drm/drm_gpuva_mgr.c + :doc: Examples + +DRM GPU VA Manager Function References +-------------------------------------- + +.. kernel-doc:: include/drm/drm_gpuva_mgr.h + :internal: + +.. kernel-doc:: drivers/gpu/drm/drm_gpuva_mgr.c + :export: + DRM Buddy Allocator ===================drivers/gpu/drm/Makefile +1
-0@@ -45,6 +45,7 @@ drm-y := \ drm_vblank.o \ drm_vblank_work.o \ drm_vma_manager.o \ + drm_gpuva_mgr.o \ drm_writeback.o drm-$(CONFIG_DRM_LEGACY) += \ drm_agpsupport.o \drivers/gpu/drm/drm_gem.c +3
-0@@ -164,6 +164,9 @@ void drm_gem_private_object_init(struct drm_device *dev,if (!obj->resv) obj->resv = &obj->_resv; +if (drm_core_check_feature (dev,DRIVER_GEM_GPUVA )) +drm_gem_gpuva_init (obj); +drm_vma_node_reset (&obj->vma_node);INIT_LIST_HEAD (&obj->lru_node); }drivers/gpu/drm/drm_gpuva_mgr.c new file +0
-0
drivers/gpu/drm/drm_gpuva_mgr.c line 2
Quoted 2 lines
drivers/gpu/drm/drm_gpuva_mgr.c new file +2
-0@@ -0,0 +1,1743 @@ +// SPDX-License-Identifier: GPL-2.0 +/*
SPDX-License is GPL-2.0 but below looks like MIT. Can we use "GPL-2.0 OR MIT" or does something restrict it to GPL-only?
Quoted 45 lines
drivers/gpu/drm/drm_gpuva_mgr.c new file +45
-0+* Copyright (c) 2022 Red Hat. +* +* Permission is hereby granted, free of charge, to any person obtaining a +* copy of this software and associated documentation files (the "Software"), +* to deal in the Software without restriction, including without limitation +* the rights to use, copy, modify, merge, publish, distribute, sublicense, +* and/or sell copies of the Software, and to permit persons to whom the +* Software is furnished to do so, subject to the following conditions: +* +* The above copyright notice and this permission notice shall be included in +* all copies or substantial portions of the Software. +* +* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL +* THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR +* OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, +* ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR +* OTHER DEALINGS IN THE SOFTWARE. +* +* Authors: +* Danilo Krummrich<[email protected]> +* +*/ + +#include <drm/drm_gpuva_mgr.h> + +#include <linux/interval_tree_generic.h> +#include <linux/mm.h> + +/** +* DOC: Overview +* +* The DRM GPU VA Manager, represented by struct drm_gpuva_manager keeps track +* of a GPU's virtual address (VA) space and manages the corresponding virtual +* mappings represented by &drm_gpuva objects. It also keeps track of the +* mapping's backing &drm_gem_object buffers. +* +* &drm_gem_object buffers maintain a list of &drm_gpuva objects representing +* all existent GPU VA mappings using this &drm_gem_object as backing buffer. +* +* GPU VAs can be flagged as sparse, such that drivers may use GPU VAs to also +* keep track of sparse PTEs in order to support Vulkan 'Sparse Resources'. +* +* The GPU VA manager internally uses a rb-tree to manage the
drivers/gpu/drm/drm_gpuva_mgr.c line 57
Quoted 10 lines
drivers/gpu/drm/drm_gpuva_mgr.c new file +10
-0+* &drm_gpuva mappings within a GPU's virtual address space. +* +* The &drm_gpuva_manager contains a special &drm_gpuva representing the +* portion of VA space reserved by the kernel. This node is initialized together +* with the GPU VA manager instance and removed when the GPU VA manager is +* destroyed. +* +* In a typical application drivers would embed struct drm_gpuva_manager and +* struct drm_gpuva within their own driver specific structures, there won't be +* any memory allocations of it's own nor memory allocations of &drm_gpuva
s/it's/its/
drivers/gpu/drm/drm_gpuva_mgr.c line 69
Quoted 12 lines
drivers/gpu/drm/drm_gpuva_mgr.c new file +12
-0+* entries. +* +* The data structures needed to store &drm_gpuvas within the &drm_gpuva_manager +* are contained within struct drm_gpuva already. Hence, for inserting +* &drm_gpuva entries from within dma-fence signalling critical sections it is +* enough to pre-allocate the &drm_gpuva structures. +*/ + +/** +* DOC: Split and Merge +* +* Besides it's capability to manage and represent a GPU VA space, the
s/it's/its/
Quoted 125 lines
drivers/gpu/drm/drm_gpuva_mgr.c new file +125
-0+* &drm_gpuva_manager also provides functions to let the &drm_gpuva_manager +* calculate a sequence of operations to satisfy a given map or unmap request. +* +* Therefore the DRM GPU VA manager provides an algorithm implementing splitting +* and merging of existent GPU VA mappings with the ones that are requested to +* be mapped or unmapped. This feature is required by the Vulkan API to +* implement Vulkan 'Sparse Memory Bindings' - drivers UAPIs often refer to this +* as VM BIND. +* +* Drivers can call drm_gpuva_sm_map() to receive a sequence of callbacks +* containing map, unmap and remap operations for a given newly requested +* mapping. The sequence of callbacks represents the set of operations to +* execute in order to integrate the new mapping cleanly into the current state +* of the GPU VA space. +* +* Depending on how the new GPU VA mapping intersects with the existent mappings +* of the GPU VA space the &drm_gpuva_fn_ops callbacks contain an arbitrary +* amount of unmap operations, a maximum of two remap operations and a single +* map operation. The caller might receive no callback at all if no operation is +* required, e.g. if the requested mapping already exists in the exact same way. +* +* The single map operation represents the original map operation requested by +* the caller. +* +* &drm_gpuva_op_unmap contains a 'keep' field, which indicates whether the +* &drm_gpuva to unmap is physically contiguous with the original mapping +* request. Optionally, if 'keep' is set, drivers may keep the actual page table +* entries for this &drm_gpuva, adding the missing page table entries only and +* update the &drm_gpuva_manager's view of things accordingly. +* +* Drivers may do the same optimization, namely delta page table updates, also +* for remap operations. This is possible since &drm_gpuva_op_remap consists of +* one unmap operation and one or two map operations, such that drivers can +* derive the page table update delta accordingly. +* +* Note that there can't be more than two existent mappings to split up, one at +* the beginning and one at the end of the new mapping, hence there is a +* maximum of two remap operations. +* +* Analogous to drm_gpuva_sm_map() drm_gpuva_sm_unmap() uses &drm_gpuva_fn_ops +* to call back into the driver in order to unmap a range of GPU VA space. The +* logic behind this function is way simpler though: For all existent mappings +* enclosed by the given range unmap operations are created. For mappings which +* are only partically located within the given range, remap operations are +* created such that those mappings are split up and re-mapped partically. +* +* As an alternative to drm_gpuva_sm_map() and drm_gpuva_sm_unmap(), +* drm_gpuva_sm_map_ops_create() and drm_gpuva_sm_unmap_ops_create() can be used +* to directly obtain an instance of struct drm_gpuva_ops containing a list of +* &drm_gpuva_op, which can be iterated with drm_gpuva_for_each_op(). This list +* contains the &drm_gpuva_ops analogous to the callbacks one would receive when +* calling drm_gpuva_sm_map() or drm_gpuva_sm_unmap(). While this way requires +* more memory (to allocate the &drm_gpuva_ops), it provides drivers a way to +* iterate the &drm_gpuva_op multiple times, e.g. once in a context where memory +* allocations are possible (e.g. to allocate GPU page tables) and once in the +* dma-fence signalling critical path. +* +* To update the &drm_gpuva_manager's view of the GPU VA space +* drm_gpuva_insert() and drm_gpuva_remove() may be used. These functions can +* safely be used from &drm_gpuva_fn_ops callbacks originating from +* drm_gpuva_sm_map() or drm_gpuva_sm_unmap(). However, it might be more +* convenient to use the provided helper functions drm_gpuva_map(), +* drm_gpuva_remap() and drm_gpuva_unmap() instead. +* +* The following diagram depicts the basic relationships of existent GPU VA +* mappings, a newly requested mapping and the resulting mappings as implemented +* by drm_gpuva_sm_map() - it doesn't cover any arbitrary combinations of these. +* +* 1) Requested mapping is identical. Replace it, but indicate the backing PTEs +* could be kept. +* +* :: +* +* 0 a 1 +* old: |-----------| (bo_offset=n) +* +* 0 a 1 +* req: |-----------| (bo_offset=n) +* +* 0 a 1 +* new: |-----------| (bo_offset=n) +* +* +* 2) Requested mapping is identical, except for the BO offset, hence replace +* the mapping. +* +* :: +* +* 0 a 1 +* old: |-----------| (bo_offset=n) +* +* 0 a 1 +* req: |-----------| (bo_offset=m) +* +* 0 a 1 +* new: |-----------| (bo_offset=m) +* +* +* 3) Requested mapping is identical, except for the backing BO, hence replace +* the mapping. +* +* :: +* +* 0 a 1 +* old: |-----------| (bo_offset=n) +* +* 0 b 1 +* req: |-----------| (bo_offset=n) +* +* 0 b 1 +* new: |-----------| (bo_offset=n) +* +* +* 4) Existent mapping is a left aligned subset of the requested one, hence +* replace the existent one. +* +* :: +* +* 0 a 1 +* old: |-----| (bo_offset=n) +* +* 0 a 2 +* req: |-----------| (bo_offset=n) +* +* 0 a 2
drivers/gpu/drm/drm_gpuva_mgr.c line 204
Quoted 10 lines
drivers/gpu/drm/drm_gpuva_mgr.c new file +10
-0+* new: |-----------| (bo_offset=n) +* +* .. note:: +* We expect to see the same result for a request with a different BO +* and/or non-contiguous BO offset. +* +* +* 5) Requested mapping's range is a left aligned subset of the existent one, +* but backed by a different BO. Hence, map the requested mapping and split +* the existent one adjusting it's BO offset.
Typo: s/it's/its/ above and in multiple places below.
Quoted 185 lines
drivers/gpu/drm/drm_gpuva_mgr.c new file +185
-0+* +* :: +* +* 0 a 2 +* old: |-----------| (bo_offset=n) +* +* 0 b 1 +* req: |-----| (bo_offset=n) +* +* 0 b 1 a' 2 +* new: |-----|-----| (b.bo_offset=n, a.bo_offset=n+1) +* +* .. note:: +* We expect to see the same result for a request with a different BO +* and/or non-contiguous BO offset. +* +* +* 6) Existent mapping is a superset of the requested mapping. Split it up, but +* indicate that the backing PTEs could be kept. +* +* :: +* +* 0 a 2 +* old: |-----------| (bo_offset=n) +* +* 0 a 1 +* req: |-----| (bo_offset=n) +* +* 0 a 1 a' 2 +* new: |-----|-----| (a.bo_offset=n, a'.bo_offset=n+1) +* +* +* 7) Requested mapping's range is a right aligned subset of the existent one, +* but backed by a different BO. Hence, map the requested mapping and split +* the existent one, without adjusting the BO offset. +* +* :: +* +* 0 a 2 +* old: |-----------| (bo_offset=n) +* +* 1 b 2 +* req: |-----| (bo_offset=m) +* +* 0 a 1 b 2 +* new: |-----|-----| (a.bo_offset=n,b.bo_offset=m) +* +* +* 8) Existent mapping is a superset of the requested mapping. Split it up, but +* indicate that the backing PTEs could be kept. +* +* :: +* +* 0 a 2 +* old: |-----------| (bo_offset=n) +* +* 1 a 2 +* req: |-----| (bo_offset=n+1) +* +* 0 a' 1 a 2 +* new: |-----|-----| (a'.bo_offset=n, a.bo_offset=n+1) +* +* +* 9) Existent mapping is overlapped at the end by the requested mapping backed +* by a different BO. Hence, map the requested mapping and split up the +* existent one, without adjusting the BO offset. +* +* :: +* +* 0 a 2 +* old: |-----------| (bo_offset=n) +* +* 1 b 3 +* req: |-----------| (bo_offset=m) +* +* 0 a 1 b 3 +* new: |-----|-----------| (a.bo_offset=n,b.bo_offset=m) +* +* +* 10) Existent mapping is overlapped by the requested mapping, both having the +* same backing BO with a contiguous offset. Indicate the backing PTEs of +* the old mapping could be kept. +* +* :: +* +* 0 a 2 +* old: |-----------| (bo_offset=n) +* +* 1 a 3 +* req: |-----------| (bo_offset=n+1) +* +* 0 a' 1 a 3 +* new: |-----|-----------| (a'.bo_offset=n, a.bo_offset=n+1) +* +* +* 11) Requested mapping's range is a centered subset of the existent one +* having a different backing BO. Hence, map the requested mapping and split +* up the existent one in two mappings, adjusting the BO offset of the right +* one accordingly. +* +* :: +* +* 0 a 3 +* old: |-----------------| (bo_offset=n) +* +* 1 b 2 +* req: |-----| (bo_offset=m) +* +* 0 a 1 b 2 a' 3 +* new: |-----|-----|-----| (a.bo_offset=n,b.bo_offset=m,a'.bo_offset=n+2) +* +* +* 12) Requested mapping is a contiguous subset of the existent one. Split it +* up, but indicate that the backing PTEs could be kept. +* +* :: +* +* 0 a 3 +* old: |-----------------| (bo_offset=n) +* +* 1 a 2 +* req: |-----| (bo_offset=n+1) +* +* 0 a' 1 a 2 a'' 3 +* old: |-----|-----|-----| (a'.bo_offset=n, a.bo_offset=n+1, a''.bo_offset=n+2) +* +* +* 13) Existent mapping is a right aligned subset of the requested one, hence +* replace the existent one. +* +* :: +* +* 1 a 2 +* old: |-----| (bo_offset=n+1) +* +* 0 a 2 +* req: |-----------| (bo_offset=n) +* +* 0 a 2 +* new: |-----------| (bo_offset=n) +* +* .. note:: +* We expect to see the same result for a request with a different bo +* and/or non-contiguous bo_offset. +* +* +* 14) Existent mapping is a centered subset of the requested one, hence +* replace the existent one. +* +* :: +* +* 1 a 2 +* old: |-----| (bo_offset=n+1) +* +* 0 a 3 +* req: |----------------| (bo_offset=n) +* +* 0 a 3 +* new: |----------------| (bo_offset=n) +* +* .. note:: +* We expect to see the same result for a request with a different bo +* and/or non-contiguous bo_offset. +* +* +* 15) Existent mappings is overlapped at the beginning by the requested mapping +* backed by a different BO. Hence, map the requested mapping and split up +* the existent one, adjusting it's BO offset accordingly. +* +* :: +* +* 1 a 3 +* old: |-----------| (bo_offset=n) +* +* 0 b 2 +* req: |-----------| (bo_offset=m) +* +* 0 b 2 a' 3 +* new: |-----------|-----| (b.bo_offset=m,a.bo_offset=n+2) +*/ + +/** +* DOC: Locking +* +* Generally, the GPU VA manager does not take care of locking itself, it is
drivers/gpu/drm/drm_gpuva_mgr.c line 399
Quoted 10 lines
drivers/gpu/drm/drm_gpuva_mgr.c new file +10
-0+* the drivers responsibility to take care about locking. Drivers might want to +* protect the following operations: inserting, removing and iterating +* &drm_gpuva objects as well as generating all kinds of operations, such as +* split / merge or prefetch. +* +* The GPU VA manager also does not take care of the locking of the backing +* &drm_gem_object buffers GPU VA lists by itself; drivers are responsible to +* enforce mutual exclusion using either the GEMs dma_resv lock or alternatively +* a driver specific external lock by setting the @DRM_GPUVA_MANAGER_LOCK_EXTERN +* flag.
Is the external lock used or anticipated by any WIP implementation? If not I suggest to leave it out until there is a user.
Quoted 201 lines
drivers/gpu/drm/drm_gpuva_mgr.c new file +201
-0+* +* For the latter, functions such as drm_gpuva_link() or drm_gpuva_unlink() +* contain lockdep checks to indicate locking issues. For this to work drivers +* must provide (in case the @DRM_GPUVA_MANAGER_LOCK_EXTERN flag is set) their +* external lock with drm_gpuva_manager_set_ext_lock() after initialization. +*/ + +/** +* DOC: Examples +* +* This section gives two examples on how to let the DRM GPUVA Manager generate +* &drm_gpuva_op in order to satisfy a given map or unmap request and how to +* make use of them. +* +* The below code is strictly limited to illustrate the generic usage pattern. +* To maintain simplicitly, it doesn't make use of any abstractions for common +* code, different (asyncronous) stages with fence signalling critical paths, +* any other helpers or error handling in terms of freeing memory and dropping +* previously taken locks. +* +* 1) Obtain a list of &drm_gpuva_op to create a new mapping:: +* +* // Allocates a new &drm_gpuva. +* struct drm_gpuva * driver_gpuva_alloc(void); +* +* // Typically drivers would embedd the &drm_gpuva_manager and &drm_gpuva +* // structure in individual driver structures and lock the dma-resv with +* // drm_exec or similar helpers. +* int driver_mapping_create(struct drm_gpuva_manager *mgr, +* u64 addr, u64 range, +* struct drm_gem_object *obj, u64 offset) +* { +* struct drm_gpuva_ops *ops; +* struct drm_gpuva_op *op +* +* driver_lock_va_space(); +* ops = drm_gpuva_sm_map_ops_create(mgr, addr, range, +* obj, offset); +* if (IS_ERR(ops)) +* return PTR_ERR(ops); +* +* drm_gpuva_for_each_op(op, ops) { +* struct drm_gpuva *va; +* +* switch (op->op) { +* case DRM_GPUVA_OP_MAP: +* va = driver_gpuva_alloc(); +* if (!va) +* ; // unwind previous VA space updates, +* // free memory and unlock +* +* driver_vm_map(); +* drm_gpuva_map(mgr, va, &op->map); +* drm_gpuva_link(va); +* +* break; +* case DRM_GPUVA_OP_REMAP: { +* struct drm_gpuva *prev = NULL, *next = NULL; +* +* va = op->remap.unmap->va; +* +* if (op->remap.prev) { +* prev = driver_gpuva_alloc(); +* if (!prev) +* ; // unwind previous VA space +* // updates, free memory and +* // unlock +* } +* +* if (op->remap.next) { +* next = driver_gpuva_alloc(); +* if (!next) +* ; // unwind previous VA space +* // updates, free memory and +* // unlock +* } +* +* driver_vm_remap(); +* drm_gpuva_remap(prev, next, &op->remap); +* +* drm_gpuva_unlink(va); +* if (prev) +* drm_gpuva_link(prev); +* if (next) +* drm_gpuva_link(next); +* +* break; +* } +* case DRM_GPUVA_OP_UNMAP: +* va = op->unmap->va; +* +* driver_vm_unmap(); +* drm_gpuva_unlink(va); +* drm_gpuva_unmap(&op->unmap); +* +* break; +* default: +* break; +* } +* } +* driver_unlock_va_space(); +* +* return 0; +* } +* +* 2) Receive a callback for each &drm_gpuva_op to create a new mapping:: +* +* struct driver_context { +* struct drm_gpuva_manager *mgr; +* struct drm_gpuva *new_va; +* struct drm_gpuva *prev_va; +* struct drm_gpuva *next_va; +* }; +* +* // ops to pass to drm_gpuva_manager_init() +* static const struct drm_gpuva_fn_ops driver_gpuva_ops = { +* .sm_step_map = driver_gpuva_map, +* .sm_step_remap = driver_gpuva_remap, +* .sm_step_unmap = driver_gpuva_unmap, +* }; +* +* // Typically drivers would embedd the &drm_gpuva_manager and &drm_gpuva +* // structure in individual driver structures and lock the dma-resv with +* // drm_exec or similar helpers. +* int driver_mapping_create(struct drm_gpuva_manager *mgr, +* u64 addr, u64 range, +* struct drm_gem_object *obj, u64 offset) +* { +* struct driver_context ctx; +* struct drm_gpuva_ops *ops; +* struct drm_gpuva_op *op; +* int ret = 0; +* +* ctx.mgr = mgr; +* +* ctx.new_va = kzalloc(sizeof(*ctx.new_va), GFP_KERNEL); +* ctx.prev_va = kzalloc(sizeof(*ctx.prev_va), GFP_KERNEL); +* ctx.next_va = kzalloc(sizeof(*ctx.next_va), GFP_KERNEL); +* if (!ctx.new_va || !ctx.prev_va || !ctx.next_va) { +* ret = -ENOMEM; +* goto out; +* } +* +* driver_lock_va_space(); +* ret = drm_gpuva_sm_map(mgr, &ctx, addr, range, obj, offset); +* driver_unlock_va_space(); +* +* out: +* kfree(ctx.new_va); +* kfree(ctx.prev_va); +* kfree(ctx.next_va); +* return ret; +* } +* +* int driver_gpuva_map(struct drm_gpuva_op *op, void *__ctx) +* { +* struct driver_context *ctx = __ctx; +* +* drm_gpuva_map(ctx->mgr, ctx->new_va, &op->map); +* +* drm_gpuva_link(ctx->new_va); +* +* // prevent the new GPUVA from being freed in +* // driver_mapping_create() +* ctx->new_va = NULL; +* +* return 0; +* } +* +* int driver_gpuva_remap(struct drm_gpuva_op *op, void *__ctx) +* { +* struct driver_context *ctx = __ctx; +* +* drm_gpuva_remap(ctx->prev_va, ctx->next_va, &op->remap); +* +* drm_gpuva_unlink(op->remap.unmap->va); +* kfree(op->remap.unmap->va); +* +* if (op->remap.prev) { +* drm_gpuva_link(ctx->prev_va); +* ctx->prev_va = NULL; +* } +* +* if (op->remap.next) { +* drm_gpuva_link(ctx->next_va); +* ctx->next_va = NULL; +* } +* +* return 0; +* } +* +* int driver_gpuva_unmap(struct drm_gpuva_op *op, void *__ctx) +* { +* drm_gpuva_unlink(op->unmap.va); +* drm_gpuva_unmap(&op->unmap); +* kfree(op->unmap.va); +* +* return 0; +* } +*/ +
drivers/gpu/drm/drm_gpuva_mgr.c line 610
Quoted 10 lines
drivers/gpu/drm/drm_gpuva_mgr.c new file +10
-0+#define to_drm_gpuva (__node) container_of((__node), struct drm_gpuva, rb.node) + +#define GPUVA_START (node) ((node)->va.addr) +#define GPUVA_LAST (node) ((node)->va.addr + (node)->va.range - 1) + +/* We do not actually use drm_gpuva_it_next(), tell the compiler to not complain +* about this. +*/ +INTERVAL_TREE_DEFINE(struct drm_gpuva, rb.node, u64, rb.__subtree_last, + GPUVA_START, GPUVA_LAST, static __attribute__((unused)),
Would s/__attribute__((unused))/__maybe_unused/ work here?
drivers/gpu/drm/drm_gpuva_mgr.c line 617
Quoted 7 lines
drivers/gpu/drm/drm_gpuva_mgr.c new file +7
-0+ drm_gpuva_it) + +static int __drm_gpuva_insert(struct drm_gpuva_manager *mgr, +struct drm_gpuva *va); +static void __drm_gpuva_remove (struct drm_gpuva *va); + +static inline bool
"static inline" is typically used only in header files, since the compiler should be smart enough to inline if beneficial. Prefer "static". Also in multiple places below.
Quoted 31 lines
drivers/gpu/drm/drm_gpuva_mgr.c new file +31
-0+drm_gpuva_check_overflow(u64 addr, u64 range) +{ +u64 end; + +return WARN (check_add_overflow (addr, range, &end), +"GPUVA address limited to %lu bytes.\n" ,sizeof (end)); +} + +static inline bool +drm_gpuva_in_mm_range (struct drm_gpuva_manager *mgr,u64 addr,u64 range) +{ +u64 end = addr + range; +u64 mm_start = mgr->mm_start; +u64 mm_end = mm_start + mgr->mm_range; + +return addr >= mm_start && end <= mm_end; +} + +static inline bool +drm_gpuva_in_kernel_node (struct drm_gpuva_manager *mgr,u64 addr,u64 range) +{ +u64 end = addr + range; +u64 kstart = mgr->kernel_alloc_node.va.addr; +u64 krange = mgr->kernel_alloc_node.va.range; +u64 kend = kstart + krange; + +return krange && addr < kend && kstart < end; +} + +static inline bool +drm_gpuva_range_valid(struct drm_gpuva_manager *mgr,
drivers/gpu/drm/drm_gpuva_mgr.c line 658
Quoted 10 lines
drivers/gpu/drm/drm_gpuva_mgr.c new file +10
-0+ u64 addr, u64 range) +{ + +return !drm_gpuva_check_overflow (addr, range) && +drm_gpuva_in_mm_range (mgr, addr, range) && + !drm_gpuva_in_kernel_node (mgr, addr, range); +} + +/** +* drm_gpuva_manager_init - initialize a &drm_gpuva_manager
Function kerneldoc names should end with "()": drm_gpuva_manager_init(). General comment across the patch.
Quoted 1,144 lines
drivers/gpu/drm/drm_gpuva_mgr.c new file +1085
-0+* @mgr: pointer to the &drm_gpuva_manager to initialize +* @name: the name of the GPU VA space +* @start_offset: the start offset of the GPU VA space +* @range: the size of the GPU VA space +* @reserve_offset: the start of the kernel reserved GPU VA area +* @reserve_range: the size of the kernel reserved GPU VA area +* @ops: &drm_gpuva_fn_ops called on &drm_gpuva_sm_map / &drm_gpuva_sm_unmap +* @flags: the feature flags for the &drm_gpuva_manager +* +* The &drm_gpuva_manager must be initialized with this function before use. +* +* Note that @mgr must be cleared to 0 before calling this function. The given +* &name is expected to be managed by the surrounding driver structures. +*/ +void +drm_gpuva_manager_init (struct drm_gpuva_manager *mgr, +const char *name, +u64 start_offset,u64 range, +u64 reserve_offset,u64 reserve_range, +const struct drm_gpuva_fn_ops *ops, +enum drm_gpuva_manager_flags flags) +{ + mgr->rb.tree =RB_ROOT_CACHED ; +INIT_LIST_HEAD (&mgr->rb.list); + +drm_gpuva_check_overflow (start_offset, range); + mgr->mm_start = start_offset; + mgr->mm_range = range; + + mgr->name = name ? name :"unknown" ; + mgr->flags = flags; + mgr->ops = ops; + +memset (&mgr->kernel_alloc_node,0 ,sizeof (struct drm_gpuva )); + +if (reserve_range) { + mgr->kernel_alloc_node.va.addr = reserve_offset; + mgr->kernel_alloc_node.va.range = reserve_range; + +if (likely (!drm_gpuva_check_overflow (reserve_offset, + reserve_range))) +__drm_gpuva_insert (mgr, &mgr->kernel_alloc_node); + } + +} +EXPORT_SYMBOL (drm_gpuva_manager_init); + +/** +* drm_gpuva_manager_destroy - cleanup a &drm_gpuva_manager +* @mgr: pointer to the &drm_gpuva_manager to clean up +* +* Note that it is a bug to call this function on a manager that still +* holds GPU VA mappings. +*/ +void +drm_gpuva_manager_destroy (struct drm_gpuva_manager *mgr) +{ + mgr->name =NULL ; + +if (mgr->kernel_alloc_node.va.range) +__drm_gpuva_remove (&mgr->kernel_alloc_node); + +WARN (!RB_EMPTY_ROOT (&mgr->rb.tree.rb_root), +"GPUVA tree is not empty, potentially leaking memory." ); +} +EXPORT_SYMBOL (drm_gpuva_manager_destroy); + +static int +__drm_gpuva_insert (struct drm_gpuva_manager *mgr, +struct drm_gpuva *va) +{ +struct rb_node *node; +struct list_head *head; + +if (drm_gpuva_it_iter_first (&mgr->rb.tree, +GPUVA_START (va), +GPUVA_LAST (va))) +return -EEXIST ; + + va->mgr = mgr; + +drm_gpuva_it_insert (va, &mgr->rb.tree); + + node =rb_prev (&va->rb.node); +if (node) + head = &(to_drm_gpuva (node))->rb.entry; +else + head = &mgr->rb.list; + +list_add (&va->rb.entry, head); + +return 0 ; +} + +/** +* drm_gpuva_insert - insert a &drm_gpuva +* @mgr: the &drm_gpuva_manager to insert the &drm_gpuva in +* @va: the &drm_gpuva to insert +* +* Insert a &drm_gpuva with a given address and range into a +* &drm_gpuva_manager. +* +* It is safe to use this function using the safe versions of iterating the GPU +* VA space, such as drm_gpuva_for_each_va_safe() and +* drm_gpuva_for_each_va_range_safe(). +* +* Returns: 0 on success, negative error code on failure. +*/ +int +drm_gpuva_insert (struct drm_gpuva_manager *mgr, +struct drm_gpuva *va) +{ +u64 addr = va->va.addr; +u64 range = va->va.range; + +if (unlikely (!drm_gpuva_range_valid (mgr, addr, range))) +return -EINVAL ; + +return __drm_gpuva_insert (mgr, va); +} +EXPORT_SYMBOL (drm_gpuva_insert); + +static void +__drm_gpuva_remove (struct drm_gpuva *va) +{ +drm_gpuva_it_remove (va, &va->mgr->rb.tree); +list_del_init (&va->rb.entry); +} + +/** +* drm_gpuva_remove - remove a &drm_gpuva +* @va: the &drm_gpuva to remove +* +* This removes the given &va from the underlaying tree. +* +* It is safe to use this function using the safe versions of iterating the GPU +* VA space, such as drm_gpuva_for_each_va_safe() and +* drm_gpuva_for_each_va_range_safe(). +*/ +void +drm_gpuva_remove (struct drm_gpuva *va) +{ +struct drm_gpuva_manager *mgr = va->mgr; + +if (unlikely (va == &mgr->kernel_alloc_node)) { +WARN (1 ,"Can't destroy kernel reserved node.\n" ); +return ; + } + +__drm_gpuva_remove (va); +} +EXPORT_SYMBOL (drm_gpuva_remove); + +/** +* drm_gpuva_link - link a &drm_gpuva +* @va: the &drm_gpuva to link +* +* This adds the given &va to the GPU VA list of the &drm_gem_object it is +* associated with. +* +* This function expects the caller to protect the GEM's GPUVA list against +* concurrent access using the GEMs dma_resv lock. +*/ +void +drm_gpuva_link (struct drm_gpuva *va) +{ +struct drm_gpuva_manager *mgr = va->mgr; +struct drm_gem_object *obj = va->gem.obj; + +if (unlikely (!obj)) +return ; + +if (drm_gpuva_manager_external_lock (mgr)) +drm_gpuva_manager_ext_assert_held (mgr); +else +dma_resv_assert_held (obj->resv); + +list_add_tail (&va->gem.entry, &obj->gpuva.list); +} +EXPORT_SYMBOL (drm_gpuva_link); + +/** +* drm_gpuva_unlink - unlink a &drm_gpuva +* @va: the &drm_gpuva to unlink +* +* This removes the given &va from the GPU VA list of the &drm_gem_object it is +* associated with. +* +* This function expects the caller to protect the GEM's GPUVA list against +* concurrent access using the GEMs dma_resv lock. +*/ +void +drm_gpuva_unlink (struct drm_gpuva *va) +{ +struct drm_gpuva_manager *mgr = va->mgr; +struct drm_gem_object *obj = va->gem.obj; + +if (unlikely (!obj)) +return ; + +if (drm_gpuva_manager_external_lock (mgr)) +drm_gpuva_manager_ext_assert_held (mgr); +else +dma_resv_assert_held (obj->resv); + +list_del_init (&va->gem.entry); +} +EXPORT_SYMBOL (drm_gpuva_unlink); + +/** +* drm_gpuva_find_first - find the first &drm_gpuva in the given range +* @mgr: the &drm_gpuva_manager to search in +* @addr: the &drm_gpuvas address +* @range: the &drm_gpuvas range +* +* Returns: the first &drm_gpuva within the given range +*/ +struct drm_gpuva * +drm_gpuva_find_first (struct drm_gpuva_manager *mgr, +u64 addr,u64 range) +{ +u64 last = addr + range -1 ; + +return drm_gpuva_it_iter_first (&mgr->rb.tree, addr, last); +} +EXPORT_SYMBOL (drm_gpuva_find_first); + +/** +* drm_gpuva_find - find a &drm_gpuva +* @mgr: the &drm_gpuva_manager to search in +* @addr: the &drm_gpuvas address +* @range: the &drm_gpuvas range +* +* Returns: the &drm_gpuva at a given &addr and with a given &range +*/ +struct drm_gpuva * +drm_gpuva_find (struct drm_gpuva_manager *mgr, +u64 addr,u64 range) +{ +struct drm_gpuva *va; + + va =drm_gpuva_find_first (mgr, addr, range); +if (!va) + goto out; + +if (va->va.addr != addr || + va->va.range != range) + goto out; + +return va; + +out: +return NULL ; +} +EXPORT_SYMBOL (drm_gpuva_find); + +/** +* drm_gpuva_find_prev - find the &drm_gpuva before the given address +* @mgr: the &drm_gpuva_manager to search in +* @start: the given GPU VA's start address +* +* Find the adjacent &drm_gpuva before the GPU VA with given &start address. +* +* Note that if there is any free space between the GPU VA mappings no mapping +* is returned. +* +* Returns: a pointer to the found &drm_gpuva or NULL if none was found +*/ +struct drm_gpuva * +drm_gpuva_find_prev (struct drm_gpuva_manager *mgr,u64 start) +{ +if (!drm_gpuva_range_valid (mgr, start -1 ,1 )) +return NULL ; + +return drm_gpuva_it_iter_first (&mgr->rb.tree, start -1 , start); +} +EXPORT_SYMBOL (drm_gpuva_find_prev); + +/** +* drm_gpuva_find_next - find the &drm_gpuva after the given address +* @mgr: the &drm_gpuva_manager to search in +* @end: the given GPU VA's end address +* +* Find the adjacent &drm_gpuva after the GPU VA with given &end address. +* +* Note that if there is any free space between the GPU VA mappings no mapping +* is returned. +* +* Returns: a pointer to the found &drm_gpuva or NULL if none was found +*/ +struct drm_gpuva * +drm_gpuva_find_next (struct drm_gpuva_manager *mgr,u64 end) +{ +if (!drm_gpuva_range_valid (mgr, end,1 )) +return NULL ; + +return drm_gpuva_it_iter_first (&mgr->rb.tree, end, end +1 ); +} +EXPORT_SYMBOL (drm_gpuva_find_next); + +/** +* drm_gpuva_interval_empty - indicate whether a given interval of the VA space +* is empty +* @mgr: the &drm_gpuva_manager to check the range for +* @addr: the start address of the range +* @range: the range of the interval +* +* Returns: true if the interval is empty, false otherwise +*/ +bool +drm_gpuva_interval_empty (struct drm_gpuva_manager *mgr,u64 addr,u64 range) +{ +return !drm_gpuva_find_first (mgr, addr, range); +} +EXPORT_SYMBOL (drm_gpuva_interval_empty); + +/** +* drm_gpuva_map - helper to insert a &drm_gpuva according to a +* &drm_gpuva_op_map +* @mgr: the &drm_gpuva_manager +* @va: the &drm_gpuva to insert +* @op: the &drm_gpuva_op_map to initialize @va with +* +* Initializes the @va from the @op and inserts it into the given @mgr. +*/ +void +drm_gpuva_map (struct drm_gpuva_manager *mgr, +struct drm_gpuva *va, +struct drm_gpuva_op_map *op) +{ +drm_gpuva_init_from_op (va, op); +drm_gpuva_insert (mgr, va); +} +EXPORT_SYMBOL (drm_gpuva_map); + +/** +* drm_gpuva_remap - helper to remap a &drm_gpuva according to a +* &drm_gpuva_op_remap +* @prev: the &drm_gpuva to remap when keeping the start of a mapping +* @next: the &drm_gpuva to remap when keeping the end of a mapping +* @op: the &drm_gpuva_op_remap to initialize @prev and @next with +* +* Removes the currently mapped &drm_gpuva and remaps it using @prev and/or +* @next. +*/ +void +drm_gpuva_remap (struct drm_gpuva *prev, +struct drm_gpuva *next, +struct drm_gpuva_op_remap *op) +{ +struct drm_gpuva *curr = op->unmap->va; +struct drm_gpuva_manager *mgr = curr->mgr; +struct drm_gpuva_op_map *map; + +drm_gpuva_remove (curr); + +if ((map = op->prev)) { +drm_gpuva_init_from_op (prev, map); +drm_gpuva_insert (mgr, prev); + } + +if ((map = op->next)) { +drm_gpuva_init_from_op (next, map); +drm_gpuva_insert (mgr, next); + } +} +EXPORT_SYMBOL (drm_gpuva_remap); + +/** +* drm_gpuva_unmap - helper to remove a &drm_gpuva according to a +* &drm_gpuva_op_unmap +* @op: the &drm_gpuva_op_unmap specifying the &drm_gpuva to remove +* +* Removes the &drm_gpuva associated with the &drm_gpuva_op_unmap. +*/ +void +drm_gpuva_unmap (struct drm_gpuva_op_unmap *op) +{ +drm_gpuva_remove (op->va); +} +EXPORT_SYMBOL (drm_gpuva_unmap); + +static int +op_map_cb (const struct drm_gpuva_fn_ops *fn,void *priv, +u64 addr,u64 range, +struct drm_gem_object *obj,u64 offset) +{ +struct drm_gpuva_op op = {}; + + op.op =DRM_GPUVA_OP_MAP ; + op.map.va.addr = addr; + op.map.va.range = range; + op.map.gem.obj = obj; + op.map.gem.offset = offset; + +return fn->sm_step_map (&op, priv); +} + +static int +op_remap_cb (const struct drm_gpuva_fn_ops *fn,void *priv, +struct drm_gpuva_op_map *prev, +struct drm_gpuva_op_map *next, +struct drm_gpuva_op_unmap *unmap) +{ +struct drm_gpuva_op op = {}; +struct drm_gpuva_op_remap *r; + + op.op =DRM_GPUVA_OP_REMAP ; + r = &op.remap; + r->prev = prev; + r->next = next; + r->unmap = unmap; + +return fn->sm_step_remap (&op, priv); +} + +static int +op_unmap_cb (const struct drm_gpuva_fn_ops *fn,void *priv, +struct drm_gpuva *va,bool merge) +{ +struct drm_gpuva_op op = {}; + + op.op =DRM_GPUVA_OP_UNMAP ; + op.unmap.va = va; + op.unmap.keep = merge; + +return fn->sm_step_unmap (&op, priv); +} + +static int +__drm_gpuva_sm_map (struct drm_gpuva_manager *mgr, +const struct drm_gpuva_fn_ops *ops,void *priv, +u64 req_addr,u64 req_range, +struct drm_gem_object *req_obj,u64 req_offset) +{ +struct drm_gpuva *va, *next, *prev =NULL ; +u64 req_end = req_addr + req_range; +int ret; + +if (unlikely (!drm_gpuva_range_valid (mgr, req_addr, req_range))) +return -EINVAL ; + +drm_gpuva_for_each_va_range_safe (va, next, mgr, req_addr, req_end) { +struct drm_gem_object *obj = va->gem.obj; +u64 offset = va->gem.offset; +u64 addr = va->va.addr; +u64 range = va->va.range; +u64 end = addr + range; +bool merge = !!va->gem.obj; + +if (addr == req_addr) { + merge &= obj == req_obj && + offset == req_offset; + +if (end == req_end) { + ret =op_unmap_cb (ops, priv, va, merge); +if (ret) +return ret; +break ; + } + +if (end < req_end) { + ret =op_unmap_cb (ops, priv, va, merge); +if (ret) +return ret; + goto next; + } + +if (end > req_end) { +struct drm_gpuva_op_map n = { + .va.addr = req_end, + .va.range = range - req_range, + .gem.obj = obj, + .gem.offset = offset + req_range, + }; +struct drm_gpuva_op_unmap u = { + .va = va, + .keep = merge, + }; + + ret =op_remap_cb (ops, priv,NULL , &n, &u); +if (ret) +return ret; +break ; + } + }else if (addr < req_addr) { +u64 ls_range = req_addr - addr; +struct drm_gpuva_op_map p = { + .va.addr = addr, + .va.range = ls_range, + .gem.obj = obj, + .gem.offset = offset, + }; +struct drm_gpuva_op_unmap u = { .va = va }; + + merge &= obj == req_obj && + offset + ls_range == req_offset; + u.keep = merge; + +if (end == req_end) { + ret =op_remap_cb (ops, priv, &p,NULL , &u); +if (ret) +return ret; +break ; + } + +if (end < req_end) { + ret =op_remap_cb (ops, priv, &p,NULL , &u); +if (ret) +return ret; + goto next; + } + +if (end > req_end) { +struct drm_gpuva_op_map n = { + .va.addr = req_end, + .va.range = end - req_end, + .gem.obj = obj, + .gem.offset = offset + ls_range + + req_range, + }; + + ret =op_remap_cb (ops, priv, &p, &n, &u); +if (ret) +return ret; +break ; + } + }else if (addr > req_addr) { + merge &= obj == req_obj && + offset == req_offset + + (addr - req_addr); + +if (end == req_end) { + ret =op_unmap_cb (ops, priv, va, merge); +if (ret) +return ret; +break ; + } + +if (end < req_end) { + ret =op_unmap_cb (ops, priv, va, merge); +if (ret) +return ret; + goto next; + } + +if (end > req_end) { +struct drm_gpuva_op_map n = { + .va.addr = req_end, + .va.range = end - req_end, + .gem.obj = obj, + .gem.offset = offset + req_end - addr, + }; +struct drm_gpuva_op_unmap u = { + .va = va, + .keep = merge, + }; + + ret =op_remap_cb (ops, priv,NULL , &n, &u); +if (ret) +return ret; +break ; + } + } +next: + prev = va; + } + +return op_map_cb (ops, priv, + req_addr, req_range, + req_obj, req_offset); +} + +static int +__drm_gpuva_sm_unmap (struct drm_gpuva_manager *mgr, +const struct drm_gpuva_fn_ops *ops,void *priv, +u64 req_addr,u64 req_range) +{ +struct drm_gpuva *va, *next; +u64 req_end = req_addr + req_range; +int ret; + +if (unlikely (!drm_gpuva_range_valid (mgr, req_addr, req_range))) +return -EINVAL ; + +drm_gpuva_for_each_va_range_safe (va, next, mgr, req_addr, req_end) { +struct drm_gpuva_op_map prev = {}, next = {}; +bool prev_split = false, next_split = false; +struct drm_gem_object *obj = va->gem.obj; +u64 offset = va->gem.offset; +u64 addr = va->va.addr; +u64 range = va->va.range; +u64 end = addr + range; + +if (addr < req_addr) { + prev.va.addr = addr; + prev.va.range = req_addr - addr; + prev.gem.obj = obj; + prev.gem.offset = offset; + + prev_split = true; + } + +if (end > req_end) { + next.va.addr = req_end; + next.va.range = end - req_end; + next.gem.obj = obj; + next.gem.offset = offset + (req_end - addr); + + next_split = true; + } + +if (prev_split || next_split) { +struct drm_gpuva_op_unmap unmap = { .va = va }; + + ret =op_remap_cb (ops, priv, + prev_split ? &prev :NULL , + next_split ? &next :NULL , + &unmap); +if (ret) +return ret; + }else { + ret =op_unmap_cb (ops, priv, va, false); +if (ret) +return ret; + } + } + +return 0 ; +} + +/** +* drm_gpuva_sm_map - creates the &drm_gpuva_op split/merge steps +* @mgr: the &drm_gpuva_manager representing the GPU VA space +* @req_addr: the start address of the new mapping +* @req_range: the range of the new mapping +* @req_obj: the &drm_gem_object to map +* @req_offset: the offset within the &drm_gem_object +* @priv: pointer to a driver private data structure +* +* This function iterates the given range of the GPU VA space. It utilizes the +* &drm_gpuva_fn_ops to call back into the driver providing the split and merge +* steps. +* +* Drivers may use these callbacks to update the GPU VA space right away within +* the callback. In case the driver decides to copy and store the operations for +* later processing neither this function nor &drm_gpuva_sm_unmap is allowed to +* be called before the &drm_gpuva_manager's view of the GPU VA space was +* updated with the previous set of operations. To update the +* &drm_gpuva_manager's view of the GPU VA space drm_gpuva_insert(), +* drm_gpuva_destroy_locked() and/or drm_gpuva_destroy_unlocked() should be +* used. +* +* A sequence of callbacks can contain map, unmap and remap operations, but +* the sequence of callbacks might also be empty if no operation is required, +* e.g. if the requested mapping already exists in the exact same way. +* +* There can be an arbitrary amount of unmap operations, a maximum of two remap +* operations and a single map operation. The latter one represents the original +* map operation requested by the caller. +* +* Returns: 0 on success or a negative error code +*/ +int +drm_gpuva_sm_map (struct drm_gpuva_manager *mgr,void *priv, +u64 req_addr,u64 req_range, +struct drm_gem_object *req_obj,u64 req_offset) +{ +const struct drm_gpuva_fn_ops *ops = mgr->ops; + +if (unlikely (!(ops && ops->sm_step_map && + ops->sm_step_remap && + ops->sm_step_unmap))) +return -EINVAL ; + +return __drm_gpuva_sm_map (mgr, ops, priv, + req_addr, req_range, + req_obj, req_offset); +} +EXPORT_SYMBOL (drm_gpuva_sm_map); + +/** +* drm_gpuva_sm_unmap - creates the &drm_gpuva_ops to split on unmap +* @mgr: the &drm_gpuva_manager representing the GPU VA space +* @priv: pointer to a driver private data structure +* @req_addr: the start address of the range to unmap +* @req_range: the range of the mappings to unmap +* +* This function iterates the given range of the GPU VA space. It utilizes the +* &drm_gpuva_fn_ops to call back into the driver providing the operations to +* unmap and, if required, split existent mappings. +* +* Drivers may use these callbacks to update the GPU VA space right away within +* the callback. In case the driver decides to copy and store the operations for +* later processing neither this function nor &drm_gpuva_sm_map is allowed to be +* called before the &drm_gpuva_manager's view of the GPU VA space was updated +* with the previous set of operations. To update the &drm_gpuva_manager's view +* of the GPU VA space drm_gpuva_insert(), drm_gpuva_destroy_locked() and/or +* drm_gpuva_destroy_unlocked() should be used. +* +* A sequence of callbacks can contain unmap and remap operations, depending on +* whether there are actual overlapping mappings to split. +* +* There can be an arbitrary amount of unmap operations and a maximum of two +* remap operations. +* +* Returns: 0 on success or a negative error code +*/ +int +drm_gpuva_sm_unmap (struct drm_gpuva_manager *mgr,void *priv, +u64 req_addr,u64 req_range) +{ +const struct drm_gpuva_fn_ops *ops = mgr->ops; + +if (unlikely (!(ops && ops->sm_step_remap && + ops->sm_step_unmap))) +return -EINVAL ; + +return __drm_gpuva_sm_unmap (mgr, ops, priv, + req_addr, req_range); +} +EXPORT_SYMBOL (drm_gpuva_sm_unmap); + +static struct drm_gpuva_op * +gpuva_op_alloc (struct drm_gpuva_manager *mgr) +{ +const struct drm_gpuva_fn_ops *fn = mgr->ops; +struct drm_gpuva_op *op; + +if (fn && fn->op_alloc) + op = fn->op_alloc (); +else + op =kzalloc (sizeof (*op),GFP_KERNEL ); + +if (unlikely (!op)) +return NULL ; + +return op; +} + +static void +gpuva_op_free (struct drm_gpuva_manager *mgr, +struct drm_gpuva_op *op) +{ +const struct drm_gpuva_fn_ops *fn = mgr->ops; + +if (fn && fn->op_free) + fn->op_free (op); +else +kfree (op); +} + +static int +drm_gpuva_sm_step (struct drm_gpuva_op *__op, +void *priv) +{ +struct { +struct drm_gpuva_manager *mgr; +struct drm_gpuva_ops *ops; + } *args = priv; +struct drm_gpuva_manager *mgr = args->mgr; +struct drm_gpuva_ops *ops = args->ops; +struct drm_gpuva_op *op; + + op =gpuva_op_alloc (mgr); +if (unlikely (!op)) + goto err; + +memcpy (op, __op,sizeof (*op)); + +if (op->op ==DRM_GPUVA_OP_REMAP ) { +struct drm_gpuva_op_remap *__r = &__op->remap; +struct drm_gpuva_op_remap *r = &op->remap; + + r->unmap =kmemdup (__r->unmap,sizeof (*r->unmap), +GFP_KERNEL ); +if (unlikely (!r->unmap)) + goto err_free_op; + +if (__r->prev) { + r->prev =kmemdup (__r->prev,sizeof (*r->prev), +GFP_KERNEL ); +if (unlikely (!r->prev)) + goto err_free_unmap; + } + +if (__r->next) { + r->next =kmemdup (__r->next,sizeof (*r->next), +GFP_KERNEL ); +if (unlikely (!r->next)) + goto err_free_prev; + } + } + +list_add_tail (&op->entry, &ops->list); + +return 0 ; + +err_free_unmap: +kfree (op->remap.unmap); +err_free_prev: +kfree (op->remap.prev); +err_free_op: +gpuva_op_free (mgr, op); +err: +return -ENOMEM ; +} + +static const struct drm_gpuva_fn_ops gpuva_list_ops = { + .sm_step_map = drm_gpuva_sm_step, + .sm_step_remap = drm_gpuva_sm_step, + .sm_step_unmap = drm_gpuva_sm_step, +}; + +/** +* drm_gpuva_sm_map_ops_create - creates the &drm_gpuva_ops to split and merge +* @mgr: the &drm_gpuva_manager representing the GPU VA space +* @req_addr: the start address of the new mapping +* @req_range: the range of the new mapping +* @req_obj: the &drm_gem_object to map +* @req_offset: the offset within the &drm_gem_object +* +* This function creates a list of operations to perform splitting and merging +* of existent mapping(s) with the newly requested one. +* +* The list can be iterated with &drm_gpuva_for_each_op and must be processed +* in the given order. It can contain map, unmap and remap operations, but it +* also can be empty if no operation is required, e.g. if the requested mapping +* already exists is the exact same way. +* +* There can be an arbitrary amount of unmap operations, a maximum of two remap +* operations and a single map operation. The latter one represents the original +* map operation requested by the caller. +* +* Note that before calling this function again with another mapping request it +* is necessary to update the &drm_gpuva_manager's view of the GPU VA space. The +* previously obtained operations must be either processed or abandoned. To +* update the &drm_gpuva_manager's view of the GPU VA space drm_gpuva_insert(), +* drm_gpuva_destroy_locked() and/or drm_gpuva_destroy_unlocked() should be +* used. +* +* After the caller finished processing the returned &drm_gpuva_ops, they must +* be freed with &drm_gpuva_ops_free. +* +* Returns: a pointer to the &drm_gpuva_ops on success, an ERR_PTR on failure +*/ +struct drm_gpuva_ops * +drm_gpuva_sm_map_ops_create (struct drm_gpuva_manager *mgr, +u64 req_addr,u64 req_range, +struct drm_gem_object *req_obj,u64 req_offset) +{ +struct drm_gpuva_ops *ops; +struct { +struct drm_gpuva_manager *mgr; +struct drm_gpuva_ops *ops; + } args; +int ret; + + ops =kzalloc (sizeof (*ops),GFP_KERNEL ); +if (unlikely (!ops)) +return ERR_PTR (-ENOMEM ); + +INIT_LIST_HEAD (&ops->list); + + args.mgr = mgr; + args.ops = ops; + + ret =__drm_gpuva_sm_map (mgr, &gpuva_list_ops, &args, + req_addr, req_range, + req_obj, req_offset); +if (ret) + goto err_free_ops; + +return ops; + +err_free_ops: +drm_gpuva_ops_free (mgr, ops); +return ERR_PTR (ret); +} +EXPORT_SYMBOL (drm_gpuva_sm_map_ops_create); + +/** +* drm_gpuva_sm_unmap_ops_create - creates the &drm_gpuva_ops to split on unmap +* @mgr: the &drm_gpuva_manager representing the GPU VA space +* @req_addr: the start address of the range to unmap +* @req_range: the range of the mappings to unmap +* +* This function creates a list of operations to perform unmapping and, if +* required, splitting of the mappings overlapping the unmap range. +* +* The list can be iterated with &drm_gpuva_for_each_op and must be processed +* in the given order. It can contain unmap and remap operations, depending on +* whether there are actual overlapping mappings to split. +* +* There can be an arbitrary amount of unmap operations and a maximum of two +* remap operations. +* +* Note that before calling this function again with another range to unmap it +* is necessary to update the &drm_gpuva_manager's view of the GPU VA space. The +* previously obtained operations must be processed or abandoned. To update the +* &drm_gpuva_manager's view of the GPU VA space drm_gpuva_insert(), +* drm_gpuva_destroy_locked() and/or drm_gpuva_destroy_unlocked() should be +* used. +* +* After the caller finished processing the returned &drm_gpuva_ops, they must +* be freed with &drm_gpuva_ops_free. +* +* Returns: a pointer to the &drm_gpuva_ops on success, an ERR_PTR on failure +*/ +struct drm_gpuva_ops * +drm_gpuva_sm_unmap_ops_create (struct drm_gpuva_manager *mgr, +u64 req_addr,u64 req_range) +{ +struct drm_gpuva_ops *ops; +struct { +struct drm_gpuva_manager *mgr; +struct drm_gpuva_ops *ops; + } args; +int ret; + + ops =kzalloc (sizeof (*ops),GFP_KERNEL ); +if (unlikely (!ops)) +return ERR_PTR (-ENOMEM ); + +INIT_LIST_HEAD (&ops->list); + + args.mgr = mgr; + args.ops = ops; + + ret =__drm_gpuva_sm_unmap (mgr, &gpuva_list_ops, &args, + req_addr, req_range); +if (ret) + goto err_free_ops; + +return ops; + +err_free_ops: +drm_gpuva_ops_free (mgr, ops); +return ERR_PTR (ret); +} +EXPORT_SYMBOL (drm_gpuva_sm_unmap_ops_create); + +/** +* drm_gpuva_prefetch_ops_create - creates the &drm_gpuva_ops to prefetch +* @mgr: the &drm_gpuva_manager representing the GPU VA space +* @addr: the start address of the range to prefetch +* @range: the range of the mappings to prefetch +* +* This function creates a list of operations to perform prefetching. +* +* The list can be iterated with &drm_gpuva_for_each_op and must be processed +* in the given order. It can contain prefetch operations. +* +* There can be an arbitrary amount of prefetch operations. +* +* After the caller finished processing the returned &drm_gpuva_ops, they must +* be freed with &drm_gpuva_ops_free. +* +* Returns: a pointer to the &drm_gpuva_ops on success, an ERR_PTR on failure +*/ +struct drm_gpuva_ops * +drm_gpuva_prefetch_ops_create (struct drm_gpuva_manager *mgr, +u64 addr,u64 range) +{ +struct drm_gpuva_ops *ops; +struct drm_gpuva_op *op; +struct drm_gpuva *va; +u64 end = addr + range; +int ret; + + ops =kzalloc (sizeof (*ops),GFP_KERNEL ); +if (!ops) +return ERR_PTR (-ENOMEM ); + +INIT_LIST_HEAD (&ops->list); + +drm_gpuva_for_each_va_range (va, mgr, addr, end) { + op =gpuva_op_alloc (mgr); +if (!op) { + ret = -ENOMEM ; + goto err_free_ops; + } + + op->op =DRM_GPUVA_OP_PREFETCH ; + op->prefetch.va = va; +list_add_tail (&op->entry, &ops->list); + } + +return ops; + +err_free_ops: +drm_gpuva_ops_free (mgr, ops); +return ERR_PTR (ret); +} +EXPORT_SYMBOL (drm_gpuva_prefetch_ops_create); + +/** +* drm_gpuva_gem_unmap_ops_create - creates the &drm_gpuva_ops to unmap a GEM +* @mgr: the &drm_gpuva_manager representing the GPU VA space +* @obj: the &drm_gem_object to unmap +* +* This function creates a list of operations to perform unmapping for every +* GPUVA attached to a GEM. +* +* The list can be iterated with &drm_gpuva_for_each_op and consists out of an +* arbitrary amount of unmap operations. +* +* After the caller finished processing the returned &drm_gpuva_ops, they must +* be freed with &drm_gpuva_ops_free. +* +* It is the callers responsibility to protect the GEMs GPUVA list against +* concurrent access using the GEMs dma_resv lock. +* +* Returns: a pointer to the &drm_gpuva_ops on success, an ERR_PTR on failure +*/ +struct drm_gpuva_ops * +drm_gpuva_gem_unmap_ops_create (struct drm_gpuva_manager *mgr, +struct drm_gem_object *obj) +{ +struct drm_gpuva_ops *ops; +struct drm_gpuva_op *op; +struct drm_gpuva *va; +int ret; + +if (drm_gpuva_manager_external_lock (mgr)) +drm_gpuva_manager_ext_assert_held (mgr); +else +dma_resv_assert_held (obj->resv); + + ops =kzalloc (sizeof (*ops),GFP_KERNEL ); +if (!ops) +return ERR_PTR (-ENOMEM ); + +INIT_LIST_HEAD (&ops->list); + +drm_gem_for_each_gpuva (va, obj) { + op =gpuva_op_alloc (mgr); +if (!op) { + ret = -ENOMEM ; + goto err_free_ops; + } + + op->op =DRM_GPUVA_OP_UNMAP ; + op->unmap.va = va; +list_add_tail (&op->entry, &ops->list); + } + +return ops; + +err_free_ops: +drm_gpuva_ops_free (mgr, ops); +return ERR_PTR (ret); +} +EXPORT_SYMBOL (drm_gpuva_gem_unmap_ops_create); + + +/** +* drm_gpuva_ops_free - free the given &drm_gpuva_ops +* @mgr: the &drm_gpuva_manager the ops were created for +* @ops: the &drm_gpuva_ops to free +* +* Frees the given &drm_gpuva_ops structure including all the ops associated +* with it. +*/ +void +drm_gpuva_ops_free (struct drm_gpuva_manager *mgr, +struct drm_gpuva_ops *ops) +{ +struct drm_gpuva_op *op, *next; + +drm_gpuva_for_each_op_safe (op, next, ops) { +list_del (&op->entry); + +if (op->op ==DRM_GPUVA_OP_REMAP ) { +kfree (op->remap.prev); +kfree (op->remap.next); +kfree (op->remap.unmap); + } + +gpuva_op_free (mgr, op); + } + +kfree (ops); +} +EXPORT_SYMBOL (drm_gpuva_ops_free);include/drm/drm_drv.h +6
-0@@ -104,6 +104,12 @@ enum drm_driver_feature {* acceleration should be handled by two drivers that are connected using auxiliary bus. */ DRIVER_COMPUTE_ACCEL =BIT (7 ), +/** +* @DRIVER_GEM_GPUVA: +* +* Driver supports user defined GPU VA bindings for GEM objects. +*/ + DRIVER_GEM_GPUVA = BIT(8),/* IMPORTANT: Below are all the legacy flags, add new ones above. */ include/drm/drm_gem.h +20
-0@@ -36,6 +36,8 @@#include <linux/kref.h> #include <linux/dma-resv.h> +#include <linux/list.h> +#include <linux/mutex.h> #include <drm/drm_vma_manager.h> @@ -379,6 +381,18 @@ struct drm_gem_object {*/ struct dma_resv _resv; +/** +* @gpuva: +* +* Provides the list of GPU VAs attached to this GEM object. +* +* Drivers should lock list accesses with the GEMs &dma_resv lock +* (&drm_gem_object.resv). +*/ +struct { +struct list_head list; + } gpuva; +/** * @funcs: * @@ -526,4 +540,42 @@ unsigned long drm_gem_lru_scan(struct drm_gem_lru *lru,int drm_gem_evict (struct drm_gem_object *obj); +/** +* drm_gem_gpuva_init - initialize the gpuva list of a GEM object +* @obj: the &drm_gem_object +* +* This initializes the &drm_gem_object's &drm_gpuva list. +*
include/drm/drm_gem.h unsigned long drm_gem_lru_scan(struct drm_gem_lru *lru, · line 558
Quoted 10 lines
include/drm/drm_gem.h +10
-0+* Calling this function is only necessary for drivers intending to support the +* &drm_driver_feature DRIVER_GEM_GPUVA. +*/ +static inline void drm_gem_gpuva_init (struct drm_gem_object *obj) +{ +INIT_LIST_HEAD (&obj->gpuva.list); +} + +/** +* drm_gem_for_each_gpuva - iternator to walk over a list of gpuvas
s/iternator/iterator/. (multiple places) Although since "iterator" typically refers to an object being iterated over, perhaps
"drm_gem_for_each_gpuva() - iterate over a list of gpuvas", (general comment across the patch).
Quoted 769 lines
include/drm/drm_gem.h +22
-0+* @entry: &drm_gpuva structure to assign to in each iteration step +* @obj: the &drm_gem_object the &drm_gpuvas to walk are associated with +* +* This iterator walks over all &drm_gpuva structures associated with the +* &drm_gpuva_manager. +*/ +#define drm_gem_for_each_gpuva (entry__, obj__) \ + list_for_each_entry(entry__, &(obj__)->gpuva.list, gem.entry) + +/** +* drm_gem_for_each_gpuva_safe - iternator to safely walk over a list of gpuvas +* @entry: &drm_gpuva structure to assign to in each iteration step +* @next: &next &drm_gpuva to store the next step +* @obj: the &drm_gem_object the &drm_gpuvas to walk are associated with +* +* This iterator walks over all &drm_gpuva structures associated with the +* &drm_gem_object. It is implemented with list_for_each_entry_safe(), hence +* it is save against removal of elements. +*/ +#define drm_gem_for_each_gpuva_safe (entry__, next__, obj__) \ + list_for_each_entry_safe(entry__, next__, &(obj__)->gpuva.list, gem.entry) +#endif /* __DRM_GEM_H__ */ include/drm/drm_gpuva_mgr.h new file +740
-0@@ -0,0 +1,756 @@ +/* SPDX-License-Identifier: GPL-2.0 */ + +#ifndef __DRM_GPUVA_MGR_H__ +#define __DRM_GPUVA_MGR_H__ + +/* +* Copyright (c) 2022 Red Hat. +* +* Permission is hereby granted, free of charge, to any person obtaining a +* copy of this software and associated documentation files (the "Software"), +* to deal in the Software without restriction, including without limitation +* the rights to use, copy, modify, merge, publish, distribute, sublicense, +* and/or sell copies of the Software, and to permit persons to whom the +* Software is furnished to do so, subject to the following conditions: +* +* The above copyright notice and this permission notice shall be included in +* all copies or substantial portions of the Software. +* +* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL +* THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR +* OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, +* ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR +* OTHER DEALINGS IN THE SOFTWARE. +*/ + +#include <linux/list.h> +#include <linux/rbtree.h> +#include <linux/types.h> + +#include <drm/drm_gem.h> + +struct drm_gpuva_manager ; +struct drm_gpuva_fn_ops ; + +/** +* enum drm_gpuva_flags - flags for struct drm_gpuva +*/ +enum drm_gpuva_flags { +/** +* @DRM_GPUVA_INVALIDATED: +* +* Flag indicating that the &drm_gpuva's backing GEM is invalidated. +*/ +DRM_GPUVA_INVALIDATED = (1 <<0 ), + +/** +* @DRM_GPUVA_SPARSE: +* +* Flag indicating that the &drm_gpuva is a sparse mapping. +*/ +DRM_GPUVA_SPARSE = (1 <<1 ), + +/** +* @DRM_GPUVA_USERBITS: user defined bits +*/ +DRM_GPUVA_USERBITS = (1 <<2 ), +}; + +/** +* struct drm_gpuva - structure to track a GPU VA mapping +* +* This structure represents a GPU VA mapping and is associated with a +* &drm_gpuva_manager. +* +* Typically, this structure is embedded in bigger driver structures. +*/ +struct drm_gpuva { +/** +* @mgr: the &drm_gpuva_manager this object is associated with +*/ +struct drm_gpuva_manager *mgr; + +/** +* @flags: the &drm_gpuva_flags for this mapping +*/ +enum drm_gpuva_flags flags; + +/** +* @va: structure containing the address and range of the &drm_gpuva +*/ +struct { +/** +* @addr: the start address +*/ +u64 addr; + +/* +* @range: the range +*/ +u64 range; + } va; + +/** +* @gem: structure containing the &drm_gem_object and it's offset +*/ +struct { +/** +* @offset: the offset within the &drm_gem_object +*/ +u64 offset; + +/** +* @obj: the mapped &drm_gem_object +*/ +struct drm_gem_object *obj; + +/** +* @entry: the &list_head to attach this object to a &drm_gem_object +*/ +struct list_head entry; + } gem; + +/** +* @rb: structure containing data to store &drm_gpuvas in a rb-tree +*/ +struct { +/** +* @rb: the rb-tree node +*/ +struct rb_node node; + +/** +* @entry: The &list_head to additionally connect &drm_gpuvas +* in the same order they appear in the interval tree. This is +* useful to keep iterating &drm_gpuvas from a start node found +* through the rb-tree while doing modifications on the rb-tree +* itself. +*/ +struct list_head entry; + +/** +* @__subtree_last: needed by the interval tree, holding last-in-subtree +*/ +u64 __subtree_last; + } rb; +}; + +int drm_gpuva_insert (struct drm_gpuva_manager *mgr,struct drm_gpuva *va); +void drm_gpuva_remove (struct drm_gpuva *va); + +void drm_gpuva_link (struct drm_gpuva *va); +void drm_gpuva_unlink (struct drm_gpuva *va); + +struct drm_gpuva *drm_gpuva_find (struct drm_gpuva_manager *mgr, +u64 addr,u64 range); +struct drm_gpuva *drm_gpuva_find_first (struct drm_gpuva_manager *mgr, +u64 addr,u64 range); +struct drm_gpuva *drm_gpuva_find_prev (struct drm_gpuva_manager *mgr,u64 start); +struct drm_gpuva *drm_gpuva_find_next (struct drm_gpuva_manager *mgr,u64 end); + +bool drm_gpuva_interval_empty (struct drm_gpuva_manager *mgr,u64 addr,u64 range); + +static inline void drm_gpuva_init (struct drm_gpuva *va,u64 addr,u64 range, +struct drm_gem_object *obj,u64 offset) +{ + va->va.addr = addr; + va->va.range = range; + va->gem.obj = obj; + va->gem.offset = offset; +} + +/** +* drm_gpuva_invalidate - sets whether the backing GEM of this &drm_gpuva is +* invalidated +* @va: the &drm_gpuva to set the invalidate flag for +* @invalidate: indicates whether the &drm_gpuva is invalidated +*/ +static inline void drm_gpuva_invalidate (struct drm_gpuva *va,bool invalidate) +{ +if (invalidate) + va->flags |=DRM_GPUVA_INVALIDATED ; +else + va->flags &= ~DRM_GPUVA_INVALIDATED ; +} + +/** +* drm_gpuva_invalidated - indicates whether the backing BO of this &drm_gpuva +* is invalidated +* @va: the &drm_gpuva to check +*/ +static inline bool drm_gpuva_invalidated (struct drm_gpuva *va) +{ +return va->flags &DRM_GPUVA_INVALIDATED ; +} + +#ifdef CONFIG_LOCKDEP +typedef struct lockdep_map *lockdep_map_p ; +#define drm_gpuva_manager_ext_assert_held (mgr) \ + lockdep_assert(lock_is_held((mgr)->ext_lock) != LOCK_STATE_NOT_HELD) +/** +* drm_gpuva_manager_set_ext_lock - set the external lock according to +* @DRM_GPUVA_MANAGER_LOCK_EXTERN +* @mgr: the &drm_gpuva_manager to set the lock for +* @lock: the lock to set +* +* If @DRM_GPUVA_MANAGER_LOCK_EXTERN is set, drivers need to call this function +* to provide the lock used to lock linking and unlinking of &drm_gpuvas to the +* &drm_gem_objects GPUVA list. +*/ +#define drm_gpuva_manager_set_ext_lock (mgr, lock) \ + (mgr)->ext_lock = &(lock)->dep_map +#else +typedef struct {/* nothing */ }lockdep_map_p ; +#define drm_gpuva_manager_ext_assert_held (mgr) do { (void)(mgr); } while (0) +#define drm_gpuva_manager_set_ext_lock (mgr, lock) do { } while (0) +#endif + +/** +* enum drm_gpuva_manager_flags - the feature flags for the &drm_gpuva_manager +*/ +enum drm_gpuva_manager_flags { +/** +* @DRM_GPUVA_MANAGER_LOCK_EXTERN: +* +* Indicates the driver has it's own external lock for linking and +* unlinking &drm_gpuvas to the &drm_gem_objects GPUVA list. +* +* When setting this flag it is rquired to set a lock via +* drm_gpuva_set_ext_lock(). +*/ +DRM_GPUVA_MANAGER_LOCK_EXTERN = (1 <<0 ), +}; + +/** +* struct drm_gpuva_manager - DRM GPU VA Manager +* +* The DRM GPU VA Manager keeps track of a GPU's virtual address space by using +* &maple_tree structures. Typically, this structure is embedded in bigger +* driver structures. +* +* Drivers can pass addresses and ranges in an arbitrary unit, e.g. bytes or +* pages. +* +* There should be one manager instance per GPU virtual address space. +*/ +struct drm_gpuva_manager { +/** +* @name: the name of the DRM GPU VA space +*/ +const char *name; + +/** +* @flags: the feature flags of the &drm_gpuva_manager +*/ +enum drm_gpuva_manager_flags flags; + +/** +* @mm_start: start of the VA space +*/ +u64 mm_start; + +/** +* @mm_range: length of the VA space +*/ +u64 mm_range; + +/** +* @rb: structures to track &drm_gpuva entries +*/ +struct { +/** +* @tree: the rb-tree to track GPU VA mappings +*/ +struct rb_root_cached tree; + +/** +* @list: the &list_head to track GPU VA mappings +*/ +struct list_head list; + } rb; + +/** +* @kernel_alloc_node: +* +* &drm_gpuva representing the address space cutout reserved for +* the kernel +*/ +struct drm_gpuva kernel_alloc_node; + +/** +* @ops: &drm_gpuva_fn_ops providing the split/merge steps to drivers +*/ +const struct drm_gpuva_fn_ops *ops; + +/** +* @ext_lock: &lockdep_map according to @DRM_GPUVA_MANAGER_LOCK_EXTERN +*/ +lockdep_map_p ext_lock; +}; + +void drm_gpuva_manager_init (struct drm_gpuva_manager *mgr, +const char *name, +u64 start_offset,u64 range, +u64 reserve_offset,u64 reserve_range, +const struct drm_gpuva_fn_ops *ops, +enum drm_gpuva_manager_flags flags); +void drm_gpuva_manager_destroy (struct drm_gpuva_manager *mgr); + +/** +* drm_gpuva_manager_external_lock - indicates whether the +* @DRM_GPUVA_MANAGER_LOCK_EXTERN flag is set +* @mgr: the &drm_gpuva_manager to check the flag for +*/ +static inline bool drm_gpuva_manager_external_lock (struct drm_gpuva_manager *mgr) +{ +return mgr->flags &DRM_GPUVA_MANAGER_LOCK_EXTERN ; +} + +/** +* drm_gpuva_for_each_va_range - iternator to walk over a range of &drm_gpuvas +* @va__: &drm_gpuva structure to assign to in each iteration step +* @mgr__: &drm_gpuva_manager to walk over +* @start__: starting offset, the first gpuva will overlap this +* @end__: ending offset, the last gpuva will start before this (but may +* overlap) +* +* This iterator walks over all &drm_gpuvas in the &drm_gpuva_manager that lie +* between @start__ and @end__. It is implemented similarly to list_for_each(), +* but is using the &drm_gpuva_manager's internal interval tree to accelerate +* the search for the starting &drm_gpuva, and hence isn't safe against removal +* of elements. It assumes that @end__ is within (or is the upper limit of) the +* &drm_gpuva_manager. This iterator does not skip over the &drm_gpuva_manager's +* @kernel_alloc_node. +*/ +#define drm_gpuva_for_each_va_range (va__, mgr__, start__, end__) \ + for (va__ = drm_gpuva_find_first((mgr__), (start__), (end__)); \ + va__ && (va__->va.addr < (end__)) && \ + !list_entry_is_head(va__, &(mgr__)->rb.list, rb.entry); \ + va__ = list_next_entry(va__, rb.entry)) + +/** +* drm_gpuva_for_each_va_range_safe - iternator to safely walk over a range of +* &drm_gpuvas +* @va__: &drm_gpuva to assign to in each iteration step +* @next__: another &drm_gpuva to use as temporary storage +* @mgr__: &drm_gpuva_manager to walk over +* @start__: starting offset, the first gpuva will overlap this +* @end__: ending offset, the last gpuva will start before this (but may +* overlap) +* +* This iterator walks over all &drm_gpuvas in the &drm_gpuva_manager that lie +* between @start__ and @end__. It is implemented similarly to +* list_for_each_safe(), but is using the &drm_gpuva_manager's internal interval +* tree to accelerate the search for the starting &drm_gpuva, and hence is safe +* against removal of elements. It assumes that @end__ is within (or is the +* upper limit of) the &drm_gpuva_manager. This iterator does not skip over the +* &drm_gpuva_manager's @kernel_alloc_node. +*/ +#define drm_gpuva_for_each_va_range_safe (va__, next__, mgr__, start__, end__) \ + for (va__ = drm_gpuva_find_first((mgr__), (start__), (end__)), \ + next__ = va ? list_next_entry(va__, rb.entry) : NULL; \ + va__ && (va__->va.addr < (end__)) && \ + !list_entry_is_head(va__, &(mgr__)->rb.list, rb.entry); \ + va__ = next__, next__ = list_next_entry(va__, rb.entry)) + +/** +* drm_gpuva_for_each_va - iternator to walk over all &drm_gpuvas +* @va__: &drm_gpuva to assign to in each iteration step +* @mgr__: &drm_gpuva_manager to walk over +* +* This iterator walks over all &drm_gpuva structures associated with the given +* &drm_gpuva_manager. +*/ +#define drm_gpuva_for_each_va (va__, mgr__) \ + list_for_each_entry(va__, &(mgr__)->rb.list, rb.entry) + +/** +* drm_gpuva_for_each_va_safe - iternator to safely walk over all &drm_gpuvas +* @va__: &drm_gpuva to assign to in each iteration step +* @next__: another &drm_gpuva to use as temporary storage +* @mgr__: &drm_gpuva_manager to walk over +* +* This iterator walks over all &drm_gpuva structures associated with the given +* &drm_gpuva_manager. It is implemented with list_for_each_entry_safe(), and +* hence safe against the removal of elements. +*/ +#define drm_gpuva_for_each_va_safe (va__, next__, mgr__) \ + list_for_each_entry_safe(va__, next__, &(mgr__)->rb.list, rb.entry) + +/** +* enum drm_gpuva_op_type - GPU VA operation type +* +* Operations to alter the GPU VA mappings tracked by the &drm_gpuva_manager. +*/ +enum drm_gpuva_op_type { +/** +* @DRM_GPUVA_OP_MAP: the map op type +*/ +DRM_GPUVA_OP_MAP , + +/** +* @DRM_GPUVA_OP_REMAP: the remap op type +*/ +DRM_GPUVA_OP_REMAP , + +/** +* @DRM_GPUVA_OP_UNMAP: the unmap op type +*/ +DRM_GPUVA_OP_UNMAP , + +/** +* @DRM_GPUVA_OP_PREFETCH: the prefetch op type +*/ +DRM_GPUVA_OP_PREFETCH , +}; + +/** +* struct drm_gpuva_op_map - GPU VA map operation +* +* This structure represents a single map operation generated by the +* DRM GPU VA manager. +*/ +struct drm_gpuva_op_map { +/** +* @va: structure containing address and range of a map +* operation +*/ +struct { +/** +* @addr: the base address of the new mapping +*/ +u64 addr; + +/** +* @range: the range of the new mapping +*/ +u64 range; + } va; + +/** +* @gem: structure containing the &drm_gem_object and it's offset +*/ +struct { +/** +* @offset: the offset within the &drm_gem_object +*/ +u64 offset; + +/** +* @obj: the &drm_gem_object to map +*/ +struct drm_gem_object *obj; + } gem; +}; + +/** +* struct drm_gpuva_op_unmap - GPU VA unmap operation +* +* This structure represents a single unmap operation generated by the +* DRM GPU VA manager. +*/ +struct drm_gpuva_op_unmap { +/** +* @va: the &drm_gpuva to unmap +*/ +struct drm_gpuva *va; + +/** +* @keep: +* +* Indicates whether this &drm_gpuva is physically contiguous with the +* original mapping request. +* +* Optionally, if &keep is set, drivers may keep the actual page table +* mappings for this &drm_gpuva, adding the missing page table entries +* only and update the &drm_gpuva_manager accordingly. +*/ +bool keep; +}; + +/** +* struct drm_gpuva_op_remap - GPU VA remap operation +* +* This represents a single remap operation generated by the DRM GPU VA manager. +* +* A remap operation is generated when an existing GPU VA mmapping is split up +* by inserting a new GPU VA mapping or by partially unmapping existent +* mapping(s), hence it consists of a maximum of two map and one unmap +* operation. +* +* The @unmap operation takes care of removing the original existing mapping. +* @prev is used to remap the preceding part, @next the subsequent part. +* +* If either a new mapping's start address is aligned with the start address +* of the old mapping or the new mapping's end address is aligned with the +* end address of the old mapping, either @prev or @next is NULL. +* +* Note, the reason for a dedicated remap operation, rather than arbitrary +* unmap and map operations, is to give drivers the chance of extracting driver +* specific data for creating the new mappings from the unmap operations's +* &drm_gpuva structure which typically is embedded in larger driver specific +* structures. +*/ +struct drm_gpuva_op_remap { +/** +* @prev: the preceding part of a split mapping +*/ +struct drm_gpuva_op_map *prev; + +/** +* @next: the subsequent part of a split mapping +*/ +struct drm_gpuva_op_map *next; + +/** +* @unmap: the unmap operation for the original existing mapping +*/ +struct drm_gpuva_op_unmap *unmap; +}; + +/** +* struct drm_gpuva_op_prefetch - GPU VA prefetch operation +* +* This structure represents a single prefetch operation generated by the +* DRM GPU VA manager. +*/ +struct drm_gpuva_op_prefetch { +/** +* @va: the &drm_gpuva to prefetch +*/ +struct drm_gpuva *va; +}; + +/** +* struct drm_gpuva_op - GPU VA operation +* +* This structure represents a single generic operation. +* +* The particular type of the operation is defined by @op. +*/ +struct drm_gpuva_op { +/** +* @entry: +* +* The &list_head used to distribute instances of this struct within +* &drm_gpuva_ops. +*/ +struct list_head entry; + +/** +* @op: the type of the operation +*/ +enum drm_gpuva_op_type op; + +union { +/** +* @map: the map operation +*/ +struct drm_gpuva_op_map map; + +/** +* @remap: the remap operation +*/ +struct drm_gpuva_op_remap remap; + +/** +* @unmap: the unmap operation +*/ +struct drm_gpuva_op_unmap unmap; + +/** +* @prefetch: the prefetch operation +*/ +struct drm_gpuva_op_prefetch prefetch; + }; +}; + +/** +* struct drm_gpuva_ops - wraps a list of &drm_gpuva_op +*/ +struct drm_gpuva_ops { +/** +* @list: the &list_head +*/ +struct list_head list; +}; + +/** +* drm_gpuva_for_each_op - iterator to walk over &drm_gpuva_ops +* @op: &drm_gpuva_op to assign in each iteration step +* @ops: &drm_gpuva_ops to walk +* +* This iterator walks over all ops within a given list of operations. +*/ +#define drm_gpuva_for_each_op (op, ops) list_for_each_entry(op, &(ops)->list, entry) + +/** +* drm_gpuva_for_each_op_safe - iterator to safely walk over &drm_gpuva_ops +* @op: &drm_gpuva_op to assign in each iteration step +* @next: &next &drm_gpuva_op to store the next step +* @ops: &drm_gpuva_ops to walk +* +* This iterator walks over all ops within a given list of operations. It is +* implemented with list_for_each_safe(), so save against removal of elements. +*/ +#define drm_gpuva_for_each_op_safe (op, next, ops) \ + list_for_each_entry_safe(op, next, &(ops)->list, entry) + +/** +* drm_gpuva_for_each_op_from_reverse - iterate backwards from the given point +* @op: &drm_gpuva_op to assign in each iteration step +* @ops: &drm_gpuva_ops to walk +* +* This iterator walks over all ops within a given list of operations beginning +* from the given operation in reverse order. +*/ +#define drm_gpuva_for_each_op_from_reverse (op, ops) \ + list_for_each_entry_from_reverse(op, &(ops)->list, entry) + +/** +* drm_gpuva_first_op - returns the first &drm_gpuva_op from &drm_gpuva_ops +* @ops: the &drm_gpuva_ops to get the fist &drm_gpuva_op from +*/ +#define drm_gpuva_first_op (ops) \ + list_first_entry(&(ops)->list, struct drm_gpuva_op, entry) + +/** +* drm_gpuva_last_op - returns the last &drm_gpuva_op from &drm_gpuva_ops +* @ops: the &drm_gpuva_ops to get the last &drm_gpuva_op from +*/ +#define drm_gpuva_last_op (ops) \ + list_last_entry(&(ops)->list, struct drm_gpuva_op, entry) + +/** +* drm_gpuva_prev_op - previous &drm_gpuva_op in the list +* @op: the current &drm_gpuva_op +*/ +#define drm_gpuva_prev_op (op) list_prev_entry(op, entry) + +/** +* drm_gpuva_next_op - next &drm_gpuva_op in the list +* @op: the current &drm_gpuva_op +*/ +#define drm_gpuva_next_op (op) list_next_entry(op, entry) + +struct drm_gpuva_ops * +drm_gpuva_sm_map_ops_create (struct drm_gpuva_manager *mgr, +u64 addr,u64 range, +struct drm_gem_object *obj,u64 offset); +struct drm_gpuva_ops * +drm_gpuva_sm_unmap_ops_create (struct drm_gpuva_manager *mgr, +u64 addr,u64 range); + +struct drm_gpuva_ops * +drm_gpuva_prefetch_ops_create (struct drm_gpuva_manager *mgr, +u64 addr,u64 range); + +struct drm_gpuva_ops * +drm_gpuva_gem_unmap_ops_create (struct drm_gpuva_manager *mgr, +struct drm_gem_object *obj); + +void drm_gpuva_ops_free (struct drm_gpuva_manager *mgr, +struct drm_gpuva_ops *ops); + +static inline void drm_gpuva_init_from_op (struct drm_gpuva *va, +struct drm_gpuva_op_map *op) +{ +drm_gpuva_init (va, op->va.addr, op->va.range, + op->gem.obj, op->gem.offset); +} + +/** +* struct drm_gpuva_fn_ops - callbacks for split/merge steps +* +* This structure defines the callbacks used by &drm_gpuva_sm_map and +* &drm_gpuva_sm_unmap to provide the split/merge steps for map and unmap +* operations to drivers. +*/ +struct drm_gpuva_fn_ops { +/** +* @op_alloc: called when the &drm_gpuva_manager allocates +* a struct drm_gpuva_op +* +* Some drivers may want to embed struct drm_gpuva_op into driver +* specific structures. By implementing this callback drivers can +* allocate memory accordingly. +* +* This callback is optional. +*/ +struct drm_gpuva_op *(*op_alloc)(void ); + +/** +* @op_free: called when the &drm_gpuva_manager frees a +* struct drm_gpuva_op +* +* Some drivers may want to embed struct drm_gpuva_op into driver +* specific structures. By implementing this callback drivers can +* free the previously allocated memory accordingly. +* +* This callback is optional. +*/ +void (*op_free)(struct drm_gpuva_op *op); + +/** +* @sm_step_map: called from &drm_gpuva_sm_map to finally insert the +* mapping once all previous steps were completed +* +* The &priv pointer matches the one the driver passed to +* &drm_gpuva_sm_map or &drm_gpuva_sm_unmap, respectively. +* +* Can be NULL if &drm_gpuva_sm_map is used. +*/ +int (*sm_step_map)(struct drm_gpuva_op *op,void *priv); + +/** +* @sm_step_remap: called from &drm_gpuva_sm_map and +* &drm_gpuva_sm_unmap to split up an existent mapping +* +* This callback is called when existent mapping needs to be split up. +* This is the case when either a newly requested mapping overlaps or +* is enclosed by an existent mapping or a partial unmap of an existent +* mapping is requested. +* +* The &priv pointer matches the one the driver passed to +* &drm_gpuva_sm_map or &drm_gpuva_sm_unmap, respectively. +* +* Can be NULL if neither &drm_gpuva_sm_map nor &drm_gpuva_sm_unmap is +* used. +*/ +int (*sm_step_remap)(struct drm_gpuva_op *op,void *priv); + +/** +* @sm_step_unmap: called from &drm_gpuva_sm_map and +* &drm_gpuva_sm_unmap to unmap an existent mapping +* +* This callback is called when existent mapping needs to be unmapped. +* This is the case when either a newly requested mapping encloses an +* existent mapping or an unmap of an existent mapping is requested. +* +* The &priv pointer matches the one the driver passed to +* &drm_gpuva_sm_map or &drm_gpuva_sm_unmap, respectively. +* +* Can be NULL if neither &drm_gpuva_sm_map nor &drm_gpuva_sm_unmap is +* used. +*/ +int (*sm_step_unmap)(struct drm_gpuva_op *op,void *priv); +}; +
include/drm/drm_gpuva_mgr.h line 750
Quoted 10 lines
include/drm/drm_gpuva_mgr.h new file +10
-0+int drm_gpuva_sm_map (struct drm_gpuva_manager *mgr,void *priv, +u64 addr,u64 range, +struct drm_gem_object *obj,u64 offset); + +int drm_gpuva_sm_unmap (struct drm_gpuva_manager *mgr,void *priv, +u64 addr,u64 range); + +void drm_gpuva_map (struct drm_gpuva_manager *mgr, +struct drm_gpuva *va, +struct drm_gpuva_op_map *op);
Missing newline
include/drm/drm_gpuva_mgr.h line 753
Quoted 3 lines
include/drm/drm_gpuva_mgr.h new file +3
-0+void drm_gpuva_remap (struct drm_gpuva *prev, +struct drm_gpuva *next, +struct drm_gpuva_op_remap *op);
Missing newline
Quoted 3 lines
include/drm/drm_gpuva_mgr.h new file +3
-0+void drm_gpuva_unmap (struct drm_gpuva_op_unmap *op); + +#endif /* __DRM_GPUVA_MGR_H__ */
Re: [PATCH drm-next v6 02/13] drm: manager to keep track of GPUs VA mappings
Hi Thomas,
Quoted 616 lines · Thomas Hellström (Intel)
Hi, Danilo
Some review comments below:
Quoted 10 lines · Danilo Krummrich
Add infrastructure to keep track of GPU virtual address (VA) mappings with a decicated VA space manager implementation.
New UAPIs, motivated by Vulkan sparse memory bindings graphics drivers start implementing, allow userspace applications to request multiple and arbitrary GPU VA mappings of buffer objects. The DRM GPU VA manager is intended to serve the following purposes in this context.
1) Provide infrastructure to track GPU VA allocations and mappings, making use of the
maple_tree.It looks like we're not using the maple tree anymore, but rather an instantiation of an interval tree.
(Perhaps as a follow-up it makes sense to provide a pre-instantiated common u64 version of the interval tree in addition to the unsigned long one since it appears to be used in multiple places in graphics drivers).
Quoted 101 lines
2) Generically connect GPU VA mappings to their backing buffers, in particular DRM GEM objects.
3) Provide a common implementation to perform more complex mapping operations on the GPU VA space. In particular splitting and merging of GPU VA mappings, e.g. for intersecting mapping requests or partial unmap requests.
Tested-by Donald Robson<[email protected]>Reviewed-by Boris Brezillon<[email protected]>Suggested-by Dave Airlie<[email protected]>Signed-off-by Danilo Krummrich<[email protected]>---
Documentation/gpu/drm-mm.rst | 36 + drivers/gpu/drm/Makefile | 1 + drivers/gpu/drm/drm_gem.c | 3 +drivers/gpu/drm/drm_gpuva_mgr.c| 1743 +++++++++++++++++++++++++++++++ include/drm/drm_drv.h | 6 + include/drm/drm_gem.h | 52 + include/drm/drm_gpuva_mgr.h | 756 ++++++++++++++7 files changed, 2597 insertions(+)create mode 100644
drivers/gpu/drm/drm_gpuva_mgr.ccreate mode 100644include/drm/drm_gpuva_mgr.hDocumentation/gpu/drm-mm.rst +0
-0@@ -466,6 +466,42 @@ DRM MM Range Allocator Function References.. kernel-doc::
drivers/gpu/drm/drm_mm.c
:export:Documentation/gpu/drm-mm.rst +36
-0+DRM GPU VA Manager +================== + +Overview +-------- + +.. kernel-doc:: drivers/gpu/drm/drm_gpuva_mgr.c + :doc: Overview + +Split and Merge +--------------- + +.. kernel-doc:: drivers/gpu/drm/drm_gpuva_mgr.c + :doc: Split and Merge + +Locking +------- + +.. kernel-doc:: drivers/gpu/drm/drm_gpuva_mgr.c + :doc: Locking + +Examples +-------- + +.. kernel-doc:: drivers/gpu/drm/drm_gpuva_mgr.c + :doc: Examples + +DRM GPU VA Manager Function References +-------------------------------------- + +.. kernel-doc:: include/drm/drm_gpuva_mgr.h + :internal: + +.. kernel-doc:: drivers/gpu/drm/drm_gpuva_mgr.c + :export: +DRM Buddy Allocator
===================drivers/gpu/drm/Makefile +1
-0@@ -45,6 +45,7 @@ drm-y := \ drm_vblank.o \ drm_vblank_work.o \ drm_vma_manager.o \ + drm_gpuva_mgr.o \drm_writeback.o
drm-$(CONFIG_DRM_LEGACY) += \
drm_agpsupport.o \drivers/gpu/drm/drm_gem.c +3
-0@@ -164,6 +164,9 @@ void drm_gem_private_object_init(struct drm_device *dev, if (!obj->resv) obj->resv = &obj->_resv; + if (drm_core_check_feature(dev, DRIVER_GEM_GPUVA)) + drm_gem_gpuva_init(obj); + drm_vma_node_reset(&obj->vma_node); INIT_LIST_HEAD(&obj->lru_node); }drivers/gpu/drm/drm_gpuva_mgr.c new file +0
-0drivers/gpu/drm/drm_gpuva_mgr.c line 2
Quoted 2 lines
drivers/gpu/drm/drm_gpuva_mgr.c new file +2
-0@@ -0,0 +1,1743 @@ +// SPDX-License-Identifier: GPL-2.0 +/* SPDX-License is GPL-2.0 but below looks like MIT. Can we use "GPL-2.0 OR MIT" or does something restrict it to GPL-only?
Quoted 68 lines
drivers/gpu/drm/drm_gpuva_mgr.c new file +3
-0+* Copyright (c) 2022 Red Hat. +* +* Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors:
+ * Danilo Krummrich<[email protected]>
+ *
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +10
-0+ +#include <drm/drm_gpuva_mgr.h> + +#include <linux/interval_tree_generic.h> +#include <linux/mm.h> + +/** +* DOC: Overview +* +* The DRM GPU VA Manager, represented by struct drm_gpuva_manager keeps track
+ * of a GPU's virtual address (VA) space and manages the corresponding virtual
+ * mappings represented by&drm_gpuvaobjects. It also keeps track of the
+ * mapping's backing&drm_gem_objectbuffers.
+ *
+ *&drm_gem_objectbuffers maintain a list of&drm_gpuvaobjects representing
+ * all existent GPU VA mappings using this&drm_gem_objectas backing buffer.
+ *
+ * GPU VAs can be flagged as sparse, such that drivers may use GPU VAs to also
+ * keep track of sparse PTEs in order to support Vulkan 'Sparse Resources'.
+ *
+ * The GPU VA manager internally uses a rb-tree to manage the
+ *&drm_gpuvamappings within a GPU's virtual address space.
+ *
+ * The&drm_gpuva_managercontains a special&drm_gpuvarepresenting the
+ * portion of VA space reserved by the kernel. This node is initialized together
+ * with the GPU VA manager instance and removed when the GPU VA manager isQuoted 8 lines
+ * destroyed.
+ *
+ * In a typical application drivers would embed structdrm_gpuva_managerand
+ * structdrm_gpuvawithin their own driver specific structures, there won't be
+ * any memory allocations of it's own nor memory allocations of&drm_gpuvas/it's/its/
Quoted 9 lines
+ * entries.
+ *
+ * The data structures needed to store&drm_gpuvaswithin the&drm_gpuva_manager
+ * are contained within structdrm_gpuvaalready. Hence, for inserting
+ *&drm_gpuvaentries from within dma-fence signalling critical sections it is
+ * enough to pre-allocate the&drm_gpuvastructures.
+ */drivers/gpu/drm/drm_gpuva_mgr.c line 69
Quoted 5 lines
drivers/gpu/drm/drm_gpuva_mgr.c new file +5
-0+ +/** +* DOC: Split and Merge +* +* Besides it's capability to manage and represent a GPU VA space, the s/it's/its/
Quoted 175 lines
+ *
&drm_gpuva_manageralso provides functions to let the&drm_gpuva_manager
+ * calculate a sequence of operations to satisfy a given map or unmap request.
+ *
+ * Therefore the DRM GPU VA manager provides an algorithm implementing splitting
+ * and merging of existent GPU VA mappings with the ones that are requested to
+ * be mapped or unmapped. This feature is required by the Vulkan API to
+ * implement Vulkan 'Sparse Memory Bindings' - drivers UAPIs often refer to this
+ * as VM BIND.
+ *
+ * Drivers can calldrm_gpuva_sm_map()to receive a sequence of callbacks
+ * containing map, unmap and remap operations for a given newly requested
+ * mapping. The sequence of callbacks represents the set of operations to
+ * execute in order to integrate the new mapping cleanly into the current state
+ * of the GPU VA space.
+ *
+ * Depending on how the new GPU VA mapping intersects with the existent mappings
+ * of the GPU VA space the&drm_gpuva_fn_opscallbacks contain an arbitrary
+ * amount of unmap operations, a maximum of two remap operations and a single
+ * map operation. The caller might receive no callback at all if no operation is
+ * required, e.g. if the requested mapping already exists in the exact same way.
+ *
+ * The single map operation represents the original map operation requested by
+ * the caller.
+ *
+ *&drm_gpuva_op_unmapcontains a 'keep' field, which indicates whether the
+ *&drm_gpuvato unmap is physically contiguous with the original mapping
+ * request. Optionally, if 'keep' is set, drivers may keep the actual page table
+ * entries for this&drm_gpuva, adding the missing page table entries only and
+ * update the &drm_gpuva_manager's view of things accordingly.
+ *
+ * Drivers may do the same optimization, namely delta page table updates, also
+ * for remap operations. This is possible since&drm_gpuva_op_remapconsists of
+ * one unmap operation and one or two map operations, such that drivers can
+ * derive the page table update delta accordingly.
+ *
+ * Note that there can't be more than two existent mappings to split up, one at
+ * the beginning and one at the end of the new mapping, hence there is a
+ * maximum of two remap operations.
+ *
+ * Analogous todrm_gpuva_sm_map()drm_gpuva_sm_unmap()uses&drm_gpuva_fn_ops
+ * to call back into the driver in order to unmap a range of GPU VA space. The
+ * logic behind this function is way simpler though: For all existent mappings
+ * enclosed by the given range unmap operations are created. For mappings which
+ * are only partically located within the given range, remap operations are
+ * created such that those mappings are split up and re-mapped partically.
+ *
+ * As an alternative todrm_gpuva_sm_map()anddrm_gpuva_sm_unmap(),
+ *drm_gpuva_sm_map_ops_create()anddrm_gpuva_sm_unmap_ops_create()can be used
+ * to directly obtain an instance of structdrm_gpuva_opscontaining a list of
+ *&drm_gpuva_op, which can be iterated withdrm_gpuva_for_each_op(). This list
+ * contains the&drm_gpuva_opsanalogous to the callbacks one would receive when
+ * callingdrm_gpuva_sm_map()ordrm_gpuva_sm_unmap(). While this way requires
+ * more memory (to allocate the&drm_gpuva_ops), it provides drivers a way to
+ * iterate the&drm_gpuva_opmultiple times, e.g. once in a context where memory
+ * allocations are possible (e.g. to allocate GPU page tables) and once in the
+ * dma-fence signalling critical path.
+ *
+ * To update the &drm_gpuva_manager's view of the GPU VA space
+ *drm_gpuva_insert()anddrm_gpuva_remove()may be used. These functions can
+ * safely be used from&drm_gpuva_fn_opscallbacks originating from
+ *drm_gpuva_sm_map()ordrm_gpuva_sm_unmap(). However, it might be more
+ * convenient to use the provided helper functionsdrm_gpuva_map(),
+ *drm_gpuva_remap()anddrm_gpuva_unmap()instead.
+ *
+ * The following diagram depicts the basic relationships of existent GPU VA
+ * mappings, a newly requested mapping and the resulting mappings as implemented
+ * bydrm_gpuva_sm_map()- it doesn't cover any arbitrary combinations of these.
+ *
+ * 1) Requested mapping is identical. Replace it, but indicate the backing PTEs
+ * could be kept.
+ *
+ * ::
+ *
+ * 0 a 1
+ * old: |-----------| (bo_offset=n)
+ *
+ * 0 a 1
+ * req: |-----------| (bo_offset=n)
+ *
+ * 0 a 1
+ * new: |-----------| (bo_offset=n)
+ *
+ *
+ * 2) Requested mapping is identical, except for the BO offset, hence replace
+ * the mapping.
+ *
+ * ::
+ *
+ * 0 a 1
+ * old: |-----------| (bo_offset=n)
+ *
+ * 0 a 1
+ * req: |-----------| (bo_offset=m)
+ *
+ * 0 a 1
+ * new: |-----------| (bo_offset=m)
+ *
+ *
+ * 3) Requested mapping is identical, except for the backing BO, hence replace
+ * the mapping.
+ *
+ * ::
+ *
+ * 0 a 1
+ * old: |-----------| (bo_offset=n)
+ *
+ * 0 b 1
+ * req: |-----------| (bo_offset=n)
+ *
+ * 0 b 1
+ * new: |-----------| (bo_offset=n)
+ *
+ *
+ * 4) Existent mapping is a left aligned subset of the requested one, hence
+ * replace the existent one.
+ *
+ * ::
+ *
+ * 0 a 1
+ * old: |-----| (bo_offset=n)
+ *
+ * 0 a 2
+ * req: |-----------| (bo_offset=n)
+ *
+ * 0 a 2
+ * new: |-----------| (bo_offset=n)
+ *
+ * .. note::
+ * We expect to see the same result for a request with a different BOQuoted 8 lines
+ * and/or non-contiguous BO offset.
+ *
+ *
+ * 5) Requested mapping's range is a left aligned subset of the existent one,
+ * but backed by a different BO. Hence, map the requested mapping and split
+ * the existent one adjusting it's BO offset.Typo: s/it's/its/ above and in multiple places below.
Quoted 216 lines
+ *
+ * ::
+ *
+ * 0 a 2
+ * old: |-----------| (bo_offset=n)
+ *
+ * 0 b 1
+ * req: |-----| (bo_offset=n)
+ *
+ * 0 b 1 a' 2
+ * new: |-----|-----| (b.bo_offset=n, a.bo_offset=n+1)
+ *
+ * .. note::
+ * We expect to see the same result for a request with a different BO
+ * and/or non-contiguous BO offset.
+ *
+ *
+ * 6) Existent mapping is a superset of the requested mapping. Split it up, but
+ * indicate that the backing PTEs could be kept.
+ *
+ * ::
+ *
+ * 0 a 2
+ * old: |-----------| (bo_offset=n)
+ *
+ * 0 a 1
+ * req: |-----| (bo_offset=n)
+ *
+ * 0 a 1 a' 2
+ * new: |-----|-----| (a.bo_offset=n, a'.bo_offset=n+1)
+ *
+ *
+ * 7) Requested mapping's range is a right aligned subset of the existent one,
+ * but backed by a different BO. Hence, map the requested mapping and split
+ * the existent one, without adjusting the BO offset.
+ *
+ * ::
+ *
+ * 0 a 2
+ * old: |-----------| (bo_offset=n)
+ *
+ * 1 b 2
+ * req: |-----| (bo_offset=m)
+ *
+ * 0 a 1 b 2
+ * new: |-----|-----| (a.bo_offset=n,b.bo_offset=m)
+ *
+ *
+ * 8) Existent mapping is a superset of the requested mapping. Split it up, but
+ * indicate that the backing PTEs could be kept.
+ *
+ * ::
+ *
+ * 0 a 2
+ * old: |-----------| (bo_offset=n)
+ *
+ * 1 a 2
+ * req: |-----| (bo_offset=n+1)
+ *
+ * 0 a' 1 a 2
+ * new: |-----|-----| (a'.bo_offset=n, a.bo_offset=n+1)
+ *
+ *
+ * 9) Existent mapping is overlapped at the end by the requested mapping backed
+ * by a different BO. Hence, map the requested mapping and split up the
+ * existent one, without adjusting the BO offset.
+ *
+ * ::
+ *
+ * 0 a 2
+ * old: |-----------| (bo_offset=n)
+ *
+ * 1 b 3
+ * req: |-----------| (bo_offset=m)
+ *
+ * 0 a 1 b 3
+ * new: |-----|-----------| (a.bo_offset=n,b.bo_offset=m)
+ *
+ *
+ * 10) Existent mapping is overlapped by the requested mapping, both having the
+ * same backing BO with a contiguous offset. Indicate the backing PTEs of
+ * the old mapping could be kept.
+ *
+ * ::
+ *
+ * 0 a 2
+ * old: |-----------| (bo_offset=n)
+ *
+ * 1 a 3
+ * req: |-----------| (bo_offset=n+1)
+ *
+ * 0 a' 1 a 3
+ * new: |-----|-----------| (a'.bo_offset=n, a.bo_offset=n+1)
+ *
+ *
+ * 11) Requested mapping's range is a centered subset of the existent one
+ * having a different backing BO. Hence, map the requested mapping and split
+ * up the existent one in two mappings, adjusting the BO offset of the right
+ * one accordingly.
+ *
+ * ::
+ *
+ * 0 a 3
+ * old: |-----------------| (bo_offset=n)
+ *
+ * 1 b 2
+ * req: |-----| (bo_offset=m)
+ *
+ * 0 a 1 b 2 a' 3
+ * new: |-----|-----|-----| (a.bo_offset=n,b.bo_offset=m,a'.bo_offset=n+2)
+ *
+ *
+ * 12) Requested mapping is a contiguous subset of the existent one. Split it
+ * up, but indicate that the backing PTEs could be kept.
+ *
+ * ::
+ *
+ * 0 a 3
+ * old: |-----------------| (bo_offset=n)
+ *
+ * 1 a 2
+ * req: |-----| (bo_offset=n+1)
+ *
+ * 0 a' 1 a 2 a'' 3
+ * old: |-----|-----|-----| (a'.bo_offset=n, a.bo_offset=n+1, a''.bo_offset=n+2)
+ *
+ *
+ * 13) Existent mapping is a right aligned subset of the requested one, hence
+ * replace the existent one.
+ *
+ * ::
+ *
+ * 1 a 2
+ * old: |-----| (bo_offset=n+1)
+ *
+ * 0 a 2
+ * req: |-----------| (bo_offset=n)
+ *
+ * 0 a 2
+ * new: |-----------| (bo_offset=n)
+ *
+ * .. note::
+ * We expect to see the same result for a request with a different bo
+ * and/or non-contiguousbo_offset.
+ *
+ *
+ * 14) Existent mapping is a centered subset of the requested one, hence
+ * replace the existent one.
+ *
+ * ::
+ *
+ * 1 a 2
+ * old: |-----| (bo_offset=n+1)
+ *
+ * 0 a 3
+ * req: |----------------| (bo_offset=n)
+ *
+ * 0 a 3
+ * new: |----------------| (bo_offset=n)
+ *
+ * .. note::
+ * We expect to see the same result for a request with a different bo
+ * and/or non-contiguousbo_offset.
+ *
+ *
+ * 15) Existent mappings is overlapped at the beginning by the requested mapping
+ * backed by a different BO. Hence, map the requested mapping and split up
+ * the existent one, adjusting it's BO offset accordingly.
+ *
+ * ::
+ *
+ * 1 a 3
+ * old: |-----------| (bo_offset=n)
+ *
+ * 0 b 2
+ * req: |-----------| (bo_offset=m)
+ *
+ * 0 b 2 a' 3
+ * new: |-----------|-----| (b.bo_offset=m,a.bo_offset=n+2)
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +5
-0+ +/** +* DOC: Locking +* +* Generally, the GPU VA manager does not take care of locking itself, it is
+ * the drivers responsibility to take care about locking. Drivers might want to
+ * protect the following operations: inserting, removing and iterating
+ *&drm_gpuvaobjects as well as generating all kinds of operations, such as
+ * split / merge or prefetch.
+ *
+ * The GPU VA manager also does not take care of the locking of the backing
+ *&drm_gem_objectbuffers GPU VA lists by itself; drivers are
Quoted 8 lines
Quoted 6 lines
responsible to
+ * enforce mutual exclusion using either the GEMsdma_resvlock or alternatively
+ * a driver specific external lock by setting the @DRM_GPUVA_MANAGER_LOCK_EXTERN
+ * flag.Is the external lock used or anticipated by any WIP implementation? If not I suggest to leave it out until there is a user.
I think the PowerVR driver requires this. However, it's generally useful for any driver using direct callbacks rather than drm_gpuva_ops.
Once the page table handling in Nouveau is re-worked, and direct callbacks can be used, I probably want to use this in Nouveau as well.
Gonna fix up all other comments.
- Danilo
Quoted 2,380 lines
Quoted 224 lines
+ *
+ * For the latter, functions such asdrm_gpuva_link()ordrm_gpuva_unlink()
+ * contain lockdep checks to indicate locking issues. For this to work drivers
+ * must provide (in case the @DRM_GPUVA_MANAGER_LOCK_EXTERN flag is set) their
+ * external lock withdrm_gpuva_manager_set_ext_lock()after initialization.
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +5
-0+ +/** +* DOC: Examples +* +* This section gives two examples on how to let the DRM GPUVA Manager generate
+ *&drm_gpuva_opin order to satisfy a given map or unmap request and how to
+ * make use of them.
+ *
+ * The below code is strictly limited to illustrate the generic usage pattern.
+ * To maintain simplicitly, it doesn't make use of any abstractions for common
+ * code, different (asyncronous) stages with fence signalling critical paths,
+ * any other helpers or error handling in terms of freeing memory and dropping
+ * previously taken locks.
+ *
+ * 1) Obtain a list of&drm_gpuva_opto create a new mapping::
+ *
+ * // Allocates a new&drm_gpuva.
+ * structdrm_gpuva*driver_gpuva_alloc(void);
+ *
+ * // Typically drivers would embedd the&drm_gpuva_managerand&drm_gpuva
+ * // structure in individual driver structures and lock the dma-resv with
+ * //drm_execor similar helpers.
+ * int driver_mapping_create(structdrm_gpuva_manager*mgr,
+ * u64 addr, u64 range,
+ * structdrm_gem_object*obj, u64 offset)
+ * {
+ * structdrm_gpuva_ops*ops;
+ * structdrm_gpuva_op*op
+ *
+ *driver_lock_va_space();
+ * ops = drm_gpuva_sm_map_ops_create(mgr, addr, range,
+ * obj, offset);
+ * if (IS_ERR(ops))
+ * returnPTR_ERR(ops);
+ *
+ *drm_gpuva_for_each_op(op, ops){
+ * structdrm_gpuva*va;
+ *
+ * switch (op->op) {
+ * caseDRM_GPUVA_OP_MAP:
+ * va =driver_gpuva_alloc();
+ * if (!va)
+ * ; // unwind previous VA space updates,
+ * // free memory and unlock
+ *
+ *driver_vm_map();
+ *drm_gpuva_map(mgr, va, &op->map);
+ *drm_gpuva_link(va);
+ *
+ * break;
+ * caseDRM_GPUVA_OP_REMAP: {
+ * structdrm_gpuva*prev = NULL, *next = NULL;
+ *
+ * va =op->remap.unmap->va;
+ *
+ * if (op->remap.prev) {
+ * prev =driver_gpuva_alloc();
+ * if (!prev)
+ * ; // unwind previous VA space
+ * // updates, free memory and
+ * // unlock
+ * }
+ *
+ * if (op->remap.next) {
+ * next =driver_gpuva_alloc();
+ * if (!next)
+ * ; // unwind previous VA space
+ * // updates, free memory and
+ * // unlock
+ * }
+ *
+ *driver_vm_remap();
+ *drm_gpuva_remap(prev, next, &op->remap);
+ *
+ *drm_gpuva_unlink(va);
+ * if (prev)
+ *drm_gpuva_link(prev);
+ * if (next)
+ *drm_gpuva_link(next);
+ *
+ * break;
+ * }
+ * caseDRM_GPUVA_OP_UNMAP:
+ * va =op->unmap->va;
+ *
+ *driver_vm_unmap();
+ *drm_gpuva_unlink(va);
+ *drm_gpuva_unmap(&op->unmap);
+ *
+ * break;
+ * default:
+ * break;
+ * }
+ * }
+ *driver_unlock_va_space();
+ *
+ * return 0;
+ * }
+ *
+ * 2) Receive a callback for each&drm_gpuva_opto create a new mapping::
+ *
+ * structdriver_context{
+ * structdrm_gpuva_manager*mgr;
+ * structdrm_gpuva*new_va;
+ * structdrm_gpuva*prev_va;
+ * structdrm_gpuva*next_va;
+ * };
+ *
+ * // ops to pass todrm_gpuva_manager_init()
+ * static const structdrm_gpuva_fn_opsdriver_gpuva_ops= {
+ * .sm_step_map =driver_gpuva_map,
+ * .sm_step_remap =driver_gpuva_remap,
+ * .sm_step_unmap =driver_gpuva_unmap,
+ * };
+ *
+ * // Typically drivers would embedd the&drm_gpuva_managerand&drm_gpuva
+ * // structure in individual driver structures and lock the dma-resv with
+ * //drm_execor similar helpers.
+ * int driver_mapping_create(structdrm_gpuva_manager*mgr,
+ * u64 addr, u64 range,
+ * structdrm_gem_object*obj, u64 offset)
+ * {
+ * structdriver_contextctx;
+ * structdrm_gpuva_ops*ops;
+ * structdrm_gpuva_op*op;
+ * int ret = 0;
+ *
+ * ctx.mgr = mgr;
+ *
+ * ctx.new_va =kzalloc(sizeof(*ctx.new_va), GFP_KERNEL);
+ * ctx.prev_va =kzalloc(sizeof(*ctx.prev_va), GFP_KERNEL);
+ * ctx.next_va =kzalloc(sizeof(*ctx.next_va), GFP_KERNEL);
+ * if (!ctx.new_va || !ctx.prev_va || !ctx.next_va) {
+ * ret =-ENOMEM;
+ * goto out;
+ * }
+ *
+ *driver_lock_va_space();
+ * ret =drm_gpuva_sm_map(mgr, &ctx, addr, range, obj, offset);
+ *driver_unlock_va_space();
+ *
+ * out:
+ *kfree(ctx.new_va);
+ *kfree(ctx.prev_va);
+ *kfree(ctx.next_va);
+ * return ret;
+ * }
+ *
+ * intdriver_gpuva_map(struct drm_gpuva_op *op, void *__ctx)
+ * {
+ * structdriver_context*ctx =__ctx;
+ *
+ *drm_gpuva_map(ctx->mgr, ctx->new_va, &op->map);
+ *
+ *drm_gpuva_link(ctx->new_va);
+ *
+ * // prevent the new GPUVA from being freed in
+ * //driver_mapping_create()
+ *ctx->new_va= NULL;
+ *
+ * return 0;
+ * }
+ *
+ * intdriver_gpuva_remap(struct drm_gpuva_op *op, void *__ctx)
+ * {
+ * structdriver_context*ctx =__ctx;
+ *
+ *drm_gpuva_remap(ctx->prev_va, ctx->next_va, &op->remap);
+ *
+ *drm_gpuva_unlink(op->remap.unmap->va);
+ *kfree(op->remap.unmap->va);
+ *
+ * if (op->remap.prev) {
+ *drm_gpuva_link(ctx->prev_va);
+ *ctx->prev_va= NULL;
+ * }
+ *
+ * if (op->remap.next) {
+ *drm_gpuva_link(ctx->next_va);
+ *ctx->next_va= NULL;
+ * }
+ *
+ * return 0;
+ * }
+ *
+ * intdriver_gpuva_unmap(struct drm_gpuva_op *op, void *__ctx)
+ * {
+ *drm_gpuva_unlink(op->unmap.va);
+ *drm_gpuva_unmap(&op->unmap);
+ *kfree(op->unmap.va);
+ *
+ * return 0;
+ * }
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +2
-0+ +#define to_drm_gpuva (__node) container_of((__node), struct
drm_gpuva, rb.node)drivers/gpu/drm/drm_gpuva_mgr.c new file +5
-0+ +#define GPUVA_START (node) ((node)->va.addr) +#define GPUVA_LAST (node) ((node)->va.addr + (node)->va.range - 1) + +/* We do not actually use drm_gpuva_it_next(), tell the compiler to Quoted 5 lines
not complain
+ * about this.
+ */
+INTERVAL_TREE_DEFINE(structdrm_gpuva, rb.node, u64, rb.__subtree_last, +GPUVA_START,GPUVA_LAST, static __attribute__((unused)),Would s/__attribute__((unused))/__maybe_unused/ work here?
drivers/gpu/drm/drm_gpuva_mgr.c line 617
Quoted 7 lines
+
drm_gpuva_it)drivers/gpu/drm/drm_gpuva_mgr.c new file +6
-0+ +static int __drm_gpuva_insert (struct drm_gpuva_manager *mgr, + struct drm_gpuva *va); +static void __drm_gpuva_remove (struct drm_gpuva *va); + +static inline bool"static inline" is typically used only in header files, since the compiler should be smart enough to inline if beneficial. Prefer "static". Also in multiple places below.
Quoted 33 lines
drivers/gpu/drm/drm_gpuva_mgr.c new file +10
-0+drm_gpuva_check_overflow(u64 addr, u64 range) +{ + u64 end; + + return WARN(check_add_overflow(addr, range, &end), + "GPUVA address limited to %lu bytes.\n", sizeof(end)); +} + +static inline bool +drm_gpuva_in_mm_range(struct drm_gpuva_manager *mgr, u64 addr, u64range)
drivers/gpu/drm/drm_gpuva_mgr.c new file +10
-0+{ + u64 end = addr + range; + u64 mm_start = mgr->mm_start; + u64 mm_end = mm_start + mgr->mm_range; + + return addr >= mm_start && end <= mm_end; +} + +static inline bool +drm_gpuva_in_kernel_node(struct drm_gpuva_manager *mgr, u64 addr, u64range)
drivers/gpu/drm/drm_gpuva_mgr.c new file +11
-0+{ + u64 end = addr + range; + u64 kstart = mgr->kernel_alloc_node.va.addr; + u64 krange = mgr->kernel_alloc_node.va.range; + u64 kend = kstart + krange; + + return krange && addr < kend && kstart < end; +} + +static inline bool +drm_gpuva_range_valid(struct drm_gpuva_manager *mgr,drivers/gpu/drm/drm_gpuva_mgr.c line 658
Quoted 10 lines
drivers/gpu/drm/drm_gpuva_mgr.c new file +10
-0+ u64 addr, u64 range) +{ + + return !drm_gpuva_check_overflow(addr, range) && + drm_gpuva_in_mm_range(mgr, addr, range) && + !drm_gpuva_in_kernel_node(mgr, addr, range); +} + +/** +* drm_gpuva_manager_init - initialize a &drm_gpuva_manager Function kerneldoc names should end with "()":
drm_gpuva_manager_init(). General comment across the patch.Quoted 1,214 lines
+ * @mgr: pointer to the
&drm_gpuva_managerto initialize
+ * @name: the name of the GPU VA space
+ * @start_offset: the start offset of the GPU VA space
+ * @range: the size of the GPU VA space
+ * @reserve_offset: the start of the kernel reserved GPU VA area
+ * @reserve_range: the size of the kernel reserved GPU VA area
+ * @ops:&drm_gpuva_fn_opscalled on&drm_gpuva_sm_map/&drm_gpuva_sm_unmap
+ * @flags: the feature flags for the&drm_gpuva_manager
+ *
+ * The&drm_gpuva_managermust be initialized with this function before use.
+ *
+ * Note that @mgr must be cleared to 0 before calling this function. The given
+ * &name is expected to be managed by the surrounding driver structures.
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +89
-0+void +drm_gpuva_manager_init (struct drm_gpuva_manager *mgr, + const char *name, + u64 start_offset, u64 range, + u64 reserve_offset, u64 reserve_range, + const struct drm_gpuva_fn_ops *ops, + enum drm_gpuva_manager_flags flags) +{ + mgr->rb.tree = RB_ROOT_CACHED; + INIT_LIST_HEAD(&mgr->rb.list); + + drm_gpuva_check_overflow(start_offset, range); + mgr->mm_start = start_offset; + mgr->mm_range = range; + + mgr->name = name ? name : "unknown"; + mgr->flags = flags; + mgr->ops = ops; + + memset(&mgr->kernel_alloc_node, 0, sizeof(struct drm_gpuva)); + + if (reserve_range) { + mgr->kernel_alloc_node.va.addr = reserve_offset; + mgr->kernel_alloc_node.va.range = reserve_range; + + if (likely(!drm_gpuva_check_overflow(reserve_offset, + reserve_range))) + __drm_gpuva_insert(mgr, &mgr->kernel_alloc_node); + } + +} +EXPORT_SYMBOL (drm_gpuva_manager_init); + +/** +* drm_gpuva_manager_destroy - cleanup a &drm_gpuva_manager +* @mgr: pointer to the &drm_gpuva_manager to clean up +* +* Note that it is a bug to call this function on a manager that still +* holds GPU VA mappings. +*/ +void +drm_gpuva_manager_destroy (struct drm_gpuva_manager *mgr) +{ + mgr->name = NULL; + + if (mgr->kernel_alloc_node.va.range) + __drm_gpuva_remove(&mgr->kernel_alloc_node); + + WARN(!RB_EMPTY_ROOT(&mgr->rb.tree.rb_root), + "GPUVA tree is not empty, potentially leaking memory."); +} +EXPORT_SYMBOL (drm_gpuva_manager_destroy); + +static int +__drm_gpuva_insert (struct drm_gpuva_manager *mgr, + struct drm_gpuva *va) +{ + struct rb_node *node; + struct list_head *head; + + if (drm_gpuva_it_iter_first(&mgr->rb.tree, + GPUVA_START(va), + GPUVA_LAST(va))) + return -EEXIST; + + va->mgr = mgr; + + drm_gpuva_it_insert(va, &mgr->rb.tree); + + node = rb_prev(&va->rb.node); + if (node) + head = &(to_drm_gpuva(node))->rb.entry; + else + head = &mgr->rb.list; + + list_add(&va->rb.entry, head); + + return 0; +} + +/** +* drm_gpuva_insert - insert a &drm_gpuva +* @mgr: the &drm_gpuva_manager to insert the &drm_gpuva in +* @va: the &drm_gpuva to insert +* +* Insert a &drm_gpuva with a given address and range into a +* &drm_gpuva_manager. +* +* It is safe to use this function using the safe versions of iterating the GPU
+ * VA space, such asdrm_gpuva_for_each_va_safe()and
+ *drm_gpuva_for_each_va_range_safe().
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +28
-0+int +drm_gpuva_insert (struct drm_gpuva_manager *mgr, + struct drm_gpuva *va) +{ + u64 addr = va->va.addr; + u64 range = va->va.range; + + if (unlikely(!drm_gpuva_range_valid(mgr, addr, range))) + return -EINVAL; + + return __drm_gpuva_insert(mgr, va); +} +EXPORT_SYMBOL (drm_gpuva_insert); + +static void +__drm_gpuva_remove (struct drm_gpuva *va) +{ + drm_gpuva_it_remove(va, &va->mgr->rb.tree); + list_del_init(&va->rb.entry); +} + +/** +* drm_gpuva_remove - remove a &drm_gpuva +* @va: the &drm_gpuva to remove +* +* This removes the given &va from the underlaying tree. +* +* It is safe to use this function using the safe versions of iterating the GPU
+ * VA space, such asdrm_gpuva_for_each_va_safe()and
+ *drm_gpuva_for_each_va_range_safe().
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +19
-0+void +drm_gpuva_remove (struct drm_gpuva *va) +{ + struct drm_gpuva_manager *mgr = va->mgr; + + if (unlikely(va == &mgr->kernel_alloc_node)) { + WARN(1, "Can't destroy kernel reserved node.\n"); + return; + } + + __drm_gpuva_remove(va); +} +EXPORT_SYMBOL (drm_gpuva_remove); + +/** +* drm_gpuva_link - link a &drm_gpuva +* @va: the &drm_gpuva to link +* +* This adds the given &va to the GPU VA list of the &drm_gem_object it is
+ * associated with.
+ *
+ * This function expects the caller to protect the GEM's GPUVA list against
+ * concurrent access using the GEMsdma_resvlock.
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +23
-0+void +drm_gpuva_link (struct drm_gpuva *va) +{ + struct drm_gpuva_manager *mgr = va->mgr; + struct drm_gem_object *obj = va->gem.obj; + + if (unlikely(!obj)) + return; + + if (drm_gpuva_manager_external_lock(mgr)) + drm_gpuva_manager_ext_assert_held(mgr); + else + dma_resv_assert_held(obj->resv); + + list_add_tail(&va->gem.entry, &obj->gpuva.list); +} +EXPORT_SYMBOL (drm_gpuva_link); + +/** +* drm_gpuva_unlink - unlink a &drm_gpuva +* @va: the &drm_gpuva to unlink +* +* This removes the given &va from the GPU VA list of the
&drm_gem_objectit is
+ * associated with.
+ *
+ * This function expects the caller to protect the GEM's GPUVA list against
+ * concurrent access using the GEMsdma_resvlock.
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +71
-0+void +drm_gpuva_unlink (struct drm_gpuva *va) +{ + struct drm_gpuva_manager *mgr = va->mgr; + struct drm_gem_object *obj = va->gem.obj; + + if (unlikely(!obj)) + return; + + if (drm_gpuva_manager_external_lock(mgr)) + drm_gpuva_manager_ext_assert_held(mgr); + else + dma_resv_assert_held(obj->resv); + + list_del_init(&va->gem.entry); +} +EXPORT_SYMBOL (drm_gpuva_unlink); + +/** +* drm_gpuva_find_first - find the first &drm_gpuva in the given range +* @mgr: the &drm_gpuva_manager to search in +* @addr: the &drm_gpuvas address +* @range: the &drm_gpuvas range +* +* Returns: the first &drm_gpuva within the given range +*/ +struct drm_gpuva * +drm_gpuva_find_first (struct drm_gpuva_manager *mgr, + u64 addr, u64 range) +{ + u64 last = addr + range - 1; + + return drm_gpuva_it_iter_first(&mgr->rb.tree, addr, last); +} +EXPORT_SYMBOL (drm_gpuva_find_first); + +/** +* drm_gpuva_find - find a &drm_gpuva +* @mgr: the &drm_gpuva_manager to search in +* @addr: the &drm_gpuvas address +* @range: the &drm_gpuvas range +* +* Returns: the &drm_gpuva at a given &addr and with a given &range +*/ +struct drm_gpuva * +drm_gpuva_find (struct drm_gpuva_manager *mgr, + u64 addr, u64 range) +{ + struct drm_gpuva *va; + + va = drm_gpuva_find_first(mgr, addr, range); + if (!va) + goto out; + + if (va->va.addr != addr || + va->va.range != range) + goto out; + + return va; + +out: + return NULL; +} +EXPORT_SYMBOL (drm_gpuva_find); + +/** +* drm_gpuva_find_prev - find the &drm_gpuva before the given address +* @mgr: the &drm_gpuva_manager to search in +* @start: the given GPU VA's start address +* +* Find the adjacent &drm_gpuva before the GPU VA with given &start address.
+ *
+ * Note that if there is any free space between the GPU VA mappings no mapping
+ * is returned.
+ *
+ * Returns: a pointer to the found&drm_gpuvaor NULL if none was found
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +16
-0+struct drm_gpuva * +drm_gpuva_find_prev (struct drm_gpuva_manager *mgr,u64 start) +{ + if (!drm_gpuva_range_valid(mgr, start - 1, 1)) + return NULL; + + return drm_gpuva_it_iter_first(&mgr->rb.tree, start - 1, start); +} +EXPORT_SYMBOL (drm_gpuva_find_prev); + +/** +* drm_gpuva_find_next - find the &drm_gpuva after the given address +* @mgr: the &drm_gpuva_manager to search in +* @end: the given GPU VA's end address +* +* Find the adjacent &drm_gpuva after the GPU VA with given &end address.
+ *
+ * Note that if there is any free space between the GPU VA mappings no mapping
+ * is returned.
+ *
+ * Returns: a pointer to the found&drm_gpuvaor NULL if none was found
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +12
-0+struct drm_gpuva * +drm_gpuva_find_next (struct drm_gpuva_manager *mgr,u64 end) +{ + if (!drm_gpuva_range_valid(mgr, end, 1)) + return NULL; + + return drm_gpuva_it_iter_first(&mgr->rb.tree, end, end + 1); +} +EXPORT_SYMBOL (drm_gpuva_find_next); + +/** +* drm_gpuva_interval_empty - indicate whether a given interval of the VA space
+ * is empty
+ * @mgr: the&drm_gpuva_managerto check the range for
+ * @addr: the start address of the range
+ * @range: the range of the interval
+ *
+ * Returns: true if the interval is empty, false otherwise
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +2
-0+bool +drm_gpuva_interval_empty(struct drm_gpuva_manager *mgr, u64 addr, u64range)
drivers/gpu/drm/drm_gpuva_mgr.c new file +32
-0+{ + return !drm_gpuva_find_first(mgr, addr, range); +} +EXPORT_SYMBOL (drm_gpuva_interval_empty); + +/** +* drm_gpuva_map - helper to insert a &drm_gpuva according to a +* &drm_gpuva_op_map +* @mgr: the &drm_gpuva_manager +* @va: the &drm_gpuva to insert +* @op: the &drm_gpuva_op_map to initialize @va with +* +* Initializes the @va from the @op and inserts it into the given @mgr. +*/ +void +drm_gpuva_map (struct drm_gpuva_manager *mgr, + struct drm_gpuva *va, + struct drm_gpuva_op_map *op) +{ + drm_gpuva_init_from_op(va, op); + drm_gpuva_insert(mgr, va); +} +EXPORT_SYMBOL (drm_gpuva_map); + +/** +* drm_gpuva_remap - helper to remap a &drm_gpuva according to a +* &drm_gpuva_op_remap +* @prev: the &drm_gpuva to remap when keeping the start of a mapping +* @next: the &drm_gpuva to remap when keeping the end of a mapping +* @op: the &drm_gpuva_op_remap to initialize @prev and @next with +* +* Removes the currently mapped &drm_gpuva and remaps it using @prev and/or
+ * @next.
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +296
-0+void +drm_gpuva_remap (struct drm_gpuva *prev, + struct drm_gpuva *next, + struct drm_gpuva_op_remap *op) +{ + struct drm_gpuva *curr = op->unmap->va; + struct drm_gpuva_manager *mgr = curr->mgr; + struct drm_gpuva_op_map *map; + + drm_gpuva_remove(curr); + + if ((map = op->prev)) { + drm_gpuva_init_from_op(prev, map); + drm_gpuva_insert(mgr, prev); + } + + if ((map = op->next)) { + drm_gpuva_init_from_op(next, map); + drm_gpuva_insert(mgr, next); + } +} +EXPORT_SYMBOL (drm_gpuva_remap); + +/** +* drm_gpuva_unmap - helper to remove a &drm_gpuva according to a +* &drm_gpuva_op_unmap +* @op: the &drm_gpuva_op_unmap specifying the &drm_gpuva to remove +* +* Removes the &drm_gpuva associated with the &drm_gpuva_op_unmap. +*/ +void +drm_gpuva_unmap (struct drm_gpuva_op_unmap *op) +{ + drm_gpuva_remove(op->va); +} +EXPORT_SYMBOL (drm_gpuva_unmap); + +static int +op_map_cb (const struct drm_gpuva_fn_ops *fn,void *priv, + u64 addr, u64 range, + struct drm_gem_object *obj, u64 offset) +{ + struct drm_gpuva_op op = {}; + + op.op = DRM_GPUVA_OP_MAP; + op.map.va.addr = addr; + op.map.va.range = range; + op.map.gem.obj = obj; + op.map.gem.offset = offset; + + return fn->sm_step_map(&op, priv); +} + +static int +op_remap_cb (const struct drm_gpuva_fn_ops *fn,void *priv, + struct drm_gpuva_op_map *prev, + struct drm_gpuva_op_map *next, + struct drm_gpuva_op_unmap *unmap) +{ + struct drm_gpuva_op op = {}; + struct drm_gpuva_op_remap *r; + + op.op = DRM_GPUVA_OP_REMAP; + r = &op.remap; + r->prev = prev; + r->next = next; + r->unmap = unmap; + + return fn->sm_step_remap(&op, priv); +} + +static int +op_unmap_cb (const struct drm_gpuva_fn_ops *fn,void *priv, + struct drm_gpuva *va, bool merge) +{ + struct drm_gpuva_op op = {}; + + op.op = DRM_GPUVA_OP_UNMAP; + op.unmap.va = va; + op.unmap.keep = merge; + + return fn->sm_step_unmap(&op, priv); +} + +static int +__drm_gpuva_sm_map (struct drm_gpuva_manager *mgr, + const struct drm_gpuva_fn_ops *ops, void *priv, + u64 req_addr, u64 req_range, + struct drm_gem_object *req_obj, u64 req_offset) +{ + struct drm_gpuva *va, *next, *prev = NULL; + u64 req_end = req_addr + req_range; + int ret; + + if (unlikely(!drm_gpuva_range_valid(mgr, req_addr, req_range))) + return -EINVAL; + + drm_gpuva_for_each_va_range_safe(va, next, mgr, req_addr, req_end) { + struct drm_gem_object *obj = va->gem.obj; + u64 offset = va->gem.offset; + u64 addr = va->va.addr; + u64 range = va->va.range; + u64 end = addr + range; + bool merge = !!va->gem.obj; + + if (addr == req_addr) { + merge &= obj == req_obj && + offset == req_offset; + + if (end == req_end) { + ret = op_unmap_cb(ops, priv, va, merge); + if (ret) + return ret; + break; + } + + if (end < req_end) { + ret = op_unmap_cb(ops, priv, va, merge); + if (ret) + return ret; + goto next; + } + + if (end > req_end) { + struct drm_gpuva_op_map n = { + .va.addr = req_end, + .va.range = range - req_range, + .gem.obj = obj, + .gem.offset = offset + req_range, + }; + struct drm_gpuva_op_unmap u = { + .va = va, + .keep = merge, + }; + + ret = op_remap_cb(ops, priv, NULL, &n, &u); + if (ret) + return ret; + break; + } + } else if (addr < req_addr) { + u64 ls_range = req_addr - addr; + struct drm_gpuva_op_map p = { + .va.addr = addr, + .va.range = ls_range, + .gem.obj = obj, + .gem.offset = offset, + }; + struct drm_gpuva_op_unmap u = { .va = va }; + + merge &= obj == req_obj && + offset + ls_range == req_offset; + u.keep = merge; + + if (end == req_end) { + ret = op_remap_cb(ops, priv, &p, NULL, &u); + if (ret) + return ret; + break; + } + + if (end < req_end) { + ret = op_remap_cb(ops, priv, &p, NULL, &u); + if (ret) + return ret; + goto next; + } + + if (end > req_end) { + struct drm_gpuva_op_map n = { + .va.addr = req_end, + .va.range = end - req_end, + .gem.obj = obj, + .gem.offset = offset + ls_range + + req_range, + }; + + ret = op_remap_cb(ops, priv, &p, &n, &u); + if (ret) + return ret; + break; + } + } else if (addr > req_addr) { + merge &= obj == req_obj && + offset == req_offset + + (addr - req_addr); + + if (end == req_end) { + ret = op_unmap_cb(ops, priv, va, merge); + if (ret) + return ret; + break; + } + + if (end < req_end) { + ret = op_unmap_cb(ops, priv, va, merge); + if (ret) + return ret; + goto next; + } + + if (end > req_end) { + struct drm_gpuva_op_map n = { + .va.addr = req_end, + .va.range = end - req_end, + .gem.obj = obj, + .gem.offset = offset + req_end - addr, + }; + struct drm_gpuva_op_unmap u = { + .va = va, + .keep = merge, + }; + + ret = op_remap_cb(ops, priv, NULL, &n, &u); + if (ret) + return ret; + break; + } + } +next: + prev = va; + } + + return op_map_cb(ops, priv, + req_addr, req_range, + req_obj, req_offset); +} + +static int +__drm_gpuva_sm_unmap (struct drm_gpuva_manager *mgr, + const struct drm_gpuva_fn_ops *ops, void *priv, + u64 req_addr, u64 req_range) +{ + struct drm_gpuva *va, *next; + u64 req_end = req_addr + req_range; + int ret; + + if (unlikely(!drm_gpuva_range_valid(mgr, req_addr, req_range))) + return -EINVAL; + + drm_gpuva_for_each_va_range_safe(va, next, mgr, req_addr, req_end) { + struct drm_gpuva_op_map prev = {}, next = {}; + bool prev_split = false, next_split = false; + struct drm_gem_object *obj = va->gem.obj; + u64 offset = va->gem.offset; + u64 addr = va->va.addr; + u64 range = va->va.range; + u64 end = addr + range; + + if (addr < req_addr) { + prev.va.addr = addr; + prev.va.range = req_addr - addr; + prev.gem.obj = obj; + prev.gem.offset = offset; + + prev_split = true; + } + + if (end > req_end) { + next.va.addr = req_end; + next.va.range = end - req_end; + next.gem.obj = obj; + next.gem.offset = offset + (req_end - addr); + + next_split = true; + } + + if (prev_split || next_split) { + struct drm_gpuva_op_unmap unmap = { .va = va }; + + ret = op_remap_cb(ops, priv, + prev_split ? &prev : NULL, + next_split ? &next : NULL, + &unmap); + if (ret) + return ret; + } else { + ret = op_unmap_cb(ops, priv, va, false); + if (ret) + return ret; + } + } + + return 0; +} + +/** +* drm_gpuva_sm_map - creates the &drm_gpuva_op split/merge steps +* @mgr: the &drm_gpuva_manager representing the GPU VA space +* @req_addr: the start address of the new mapping +* @req_range: the range of the new mapping +* @req_obj: the &drm_gem_object to map +* @req_offset: the offset within the &drm_gem_object +* @priv: pointer to a driver private data structure +* +* This function iterates the given range of the GPU VA space. It utilizes the
+ *&drm_gpuva_fn_opsto call back into the driver providing the split and merge
+ * steps.
+ *
+ * Drivers may use these callbacks to update the GPU VA space right away within
+ * the callback. In case the driver decides to copy and store the operations for
+ * later processing neither this function nor&drm_gpuva_sm_unmapis allowed to
+ * be called before the &drm_gpuva_manager's view of the GPU VA space was
+ * updated with the previous set of operations. To update the
+ * &drm_gpuva_manager's view of the GPU VA spacedrm_gpuva_insert(),
+ *drm_gpuva_destroy_locked()and/ordrm_gpuva_destroy_unlocked()should be
+ * used.
+ *
+ * A sequence of callbacks can contain map, unmap and remap operations, but
+ * the sequence of callbacks might also be empty if no operation is required,
+ * e.g. if the requested mapping already exists in the exact same way.
+ *
+ * There can be an arbitrary amount of unmap operations, a maximum of two remap
+ * operations and a single map operation. The latter one represents the original
+ * map operation requested by the caller.
+ *
+ * Returns: 0 on success or a negative error code
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +26
-0+int +drm_gpuva_sm_map (struct drm_gpuva_manager *mgr,void *priv, + u64 req_addr, u64 req_range, + struct drm_gem_object *req_obj, u64 req_offset) +{ + const struct drm_gpuva_fn_ops *ops = mgr->ops; + + if (unlikely(!(ops && ops->sm_step_map && + ops->sm_step_remap && + ops->sm_step_unmap))) + return -EINVAL; + + return __drm_gpuva_sm_map(mgr, ops, priv, + req_addr, req_range, + req_obj, req_offset); +} +EXPORT_SYMBOL (drm_gpuva_sm_map); + +/** +* drm_gpuva_sm_unmap - creates the &drm_gpuva_ops to split on unmap +* @mgr: the &drm_gpuva_manager representing the GPU VA space +* @priv: pointer to a driver private data structure +* @req_addr: the start address of the range to unmap +* @req_range: the range of the mappings to unmap +* +* This function iterates the given range of the GPU VA space. It utilizes the
+ *&drm_gpuva_fn_opsto call back into the driver providing the operations to
+ * unmap and, if required, split existent mappings.
+ *
+ * Drivers may use these callbacks to update the GPU VA space right away within
+ * the callback. In case the driver decides to copy and store the operations for
+ * later processing neither this function nor&drm_gpuva_sm_mapis allowed to be
+ * called before the &drm_gpuva_manager's view of the GPU VA space was updated
+ * with the previous set of operations. To update the &drm_gpuva_manager's view
+ * of the GPU VA spacedrm_gpuva_insert(),drm_gpuva_destroy_locked()and/or
+ *drm_gpuva_destroy_unlocked()should be used.
+ *
+ * A sequence of callbacks can contain unmap and remap operations, depending on
+ * whether there are actual overlapping mappings to split.
+ *
+ * There can be an arbitrary amount of unmap operations and a maximum of two
+ * remap operations.
+ *
+ * Returns: 0 on success or a negative error code
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +108
-0+int +drm_gpuva_sm_unmap (struct drm_gpuva_manager *mgr,void *priv, + u64 req_addr, u64 req_range) +{ + const struct drm_gpuva_fn_ops *ops = mgr->ops; + + if (unlikely(!(ops && ops->sm_step_remap && + ops->sm_step_unmap))) + return -EINVAL; + + return __drm_gpuva_sm_unmap(mgr, ops, priv, + req_addr, req_range); +} +EXPORT_SYMBOL (drm_gpuva_sm_unmap); + +static struct drm_gpuva_op * +gpuva_op_alloc (struct drm_gpuva_manager *mgr) +{ + const struct drm_gpuva_fn_ops *fn = mgr->ops; + struct drm_gpuva_op *op; + + if (fn && fn->op_alloc) + op = fn->op_alloc(); + else + op = kzalloc(sizeof(*op), GFP_KERNEL); + + if (unlikely(!op)) + return NULL; + + return op; +} + +static void +gpuva_op_free (struct drm_gpuva_manager *mgr, + struct drm_gpuva_op *op) +{ + const struct drm_gpuva_fn_ops *fn = mgr->ops; + + if (fn && fn->op_free) + fn->op_free(op); + else + kfree(op); +} + +static int +drm_gpuva_sm_step (struct drm_gpuva_op *__op, + void *priv) +{ + struct { + struct drm_gpuva_manager *mgr; + struct drm_gpuva_ops *ops; + } *args = priv; + struct drm_gpuva_manager *mgr = args->mgr; + struct drm_gpuva_ops *ops = args->ops; + struct drm_gpuva_op *op; + + op = gpuva_op_alloc(mgr); + if (unlikely(!op)) + goto err; + + memcpy(op, __op, sizeof(*op)); + + if (op->op == DRM_GPUVA_OP_REMAP) { + struct drm_gpuva_op_remap *__r = &__op->remap; + struct drm_gpuva_op_remap *r = &op->remap; + + r->unmap = kmemdup(__r->unmap, sizeof(*r->unmap), + GFP_KERNEL); + if (unlikely(!r->unmap)) + goto err_free_op; + + if (__r->prev) { + r->prev = kmemdup(__r->prev, sizeof(*r->prev), + GFP_KERNEL); + if (unlikely(!r->prev)) + goto err_free_unmap; + } + + if (__r->next) { + r->next = kmemdup(__r->next, sizeof(*r->next), + GFP_KERNEL); + if (unlikely(!r->next)) + goto err_free_prev; + } + } + + list_add_tail(&op->entry, &ops->list); + + return 0; + +err_free_unmap: + kfree(op->remap.unmap); +err_free_prev: + kfree(op->remap.prev); +err_free_op: + gpuva_op_free(mgr, op); +err: + return -ENOMEM; +} + +static const struct drm_gpuva_fn_ops gpuva_list_ops = { + .sm_step_map = drm_gpuva_sm_step, + .sm_step_remap = drm_gpuva_sm_step, + .sm_step_unmap = drm_gpuva_sm_step, +}; + +/** +* drm_gpuva_sm_map_ops_create - creates the &drm_gpuva_ops to split and merge
+ * @mgr: the&drm_gpuva_managerrepresenting the GPU VA space
+ * @req_addr: the start address of the new mapping
+ * @req_range: the range of the new mapping
+ * @req_obj: the&drm_gem_objectto map
+ * @req_offset: the offset within the&drm_gem_object
+ *
+ * This function creates a list of operations to perform splitting and merging
+ * of existent mapping(s) with the newly requested one.
+ *
+ * The list can be iterated with&drm_gpuva_for_each_opand must be processed
+ * in the given order. It can contain map, unmap and remap operations, but it
+ * also can be empty if no operation is required, e.g. if the requested mapping
+ * already exists is the exact same way.
+ *
+ * There can be an arbitrary amount of unmap operations, a maximum of two remap
+ * operations and a single map operation. The latter one represents the original
+ * map operation requested by the caller.
+ *
+ * Note that before calling this function again with another mapping request it
+ * is necessary to update the &drm_gpuva_manager's view of the GPU VA space. The
+ * previously obtained operations must be either processed or abandoned. To
+ * update the &drm_gpuva_manager's view of the GPU VA spacedrm_gpuva_insert(),
+ *drm_gpuva_destroy_locked()and/ordrm_gpuva_destroy_unlocked()should be
+ * used.
+ *
+ * After the caller finished processing the returned&drm_gpuva_ops, they must
+ * be freed with&drm_gpuva_ops_free.
+ *
+ * Returns: a pointer to the&drm_gpuva_opson success, anERR_PTRon failure
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +37
-0+struct drm_gpuva_ops * +drm_gpuva_sm_map_ops_create (struct drm_gpuva_manager *mgr, + u64 req_addr, u64 req_range, + struct drm_gem_object *req_obj, u64 req_offset) +{ + struct drm_gpuva_ops *ops; + struct { + struct drm_gpuva_manager *mgr; + struct drm_gpuva_ops *ops; + } args; + int ret; + + ops = kzalloc(sizeof(*ops), GFP_KERNEL); + if (unlikely(!ops)) + return ERR_PTR(-ENOMEM); + + INIT_LIST_HEAD(&ops->list); + + args.mgr = mgr; + args.ops = ops; + + ret = __drm_gpuva_sm_map(mgr, &gpuva_list_ops, &args, + req_addr, req_range, + req_obj, req_offset); + if (ret) + goto err_free_ops; + + return ops; + +err_free_ops: + drm_gpuva_ops_free(mgr, ops); + return ERR_PTR(ret); +} +EXPORT_SYMBOL (drm_gpuva_sm_map_ops_create); + +/** +* drm_gpuva_sm_unmap_ops_create - creates the &drm_gpuva_ops to split on unmap
+ * @mgr: the&drm_gpuva_managerrepresenting the GPU VA space
+ * @req_addr: the start address of the range to unmap
+ * @req_range: the range of the mappings to unmap
+ *
+ * This function creates a list of operations to perform unmapping and, if
+ * required, splitting of the mappings overlapping the unmap range.
+ *
+ * The list can be iterated with&drm_gpuva_for_each_opand must be processed
+ * in the given order. It can contain unmap and remap operations, depending on
+ * whether there are actual overlapping mappings to split.
+ *
+ * There can be an arbitrary amount of unmap operations and a maximum of two
+ * remap operations.
+ *
+ * Note that before calling this function again with another range to unmap it
+ * is necessary to update the &drm_gpuva_manager's view of the GPU VA space. The
+ * previously obtained operations must be processed or abandoned. To update the
+ * &drm_gpuva_manager's view of the GPU VA spacedrm_gpuva_insert(),
+ *drm_gpuva_destroy_locked()and/ordrm_gpuva_destroy_unlocked()should be
+ * used.
+ *
+ * After the caller finished processing the returned&drm_gpuva_ops, they must
+ * be freed with&drm_gpuva_ops_free.
+ *
+ * Returns: a pointer to the&drm_gpuva_opson success, anERR_PTRon failure
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +35
-0+struct drm_gpuva_ops * +drm_gpuva_sm_unmap_ops_create (struct drm_gpuva_manager *mgr, + u64 req_addr, u64 req_range) +{ + struct drm_gpuva_ops *ops; + struct { + struct drm_gpuva_manager *mgr; + struct drm_gpuva_ops *ops; + } args; + int ret; + + ops = kzalloc(sizeof(*ops), GFP_KERNEL); + if (unlikely(!ops)) + return ERR_PTR(-ENOMEM); + + INIT_LIST_HEAD(&ops->list); + + args.mgr = mgr; + args.ops = ops; + + ret = __drm_gpuva_sm_unmap(mgr, &gpuva_list_ops, &args, + req_addr, req_range); + if (ret) + goto err_free_ops; + + return ops; + +err_free_ops: + drm_gpuva_ops_free(mgr, ops); + return ERR_PTR(ret); +} +EXPORT_SYMBOL (drm_gpuva_sm_unmap_ops_create); + +/** +* drm_gpuva_prefetch_ops_create - creates the &drm_gpuva_ops to prefetch
+ * @mgr: the&drm_gpuva_managerrepresenting the GPU VA space
+ * @addr: the start address of the range to prefetch
+ * @range: the range of the mappings to prefetch
+ *
+ * This function creates a list of operations to perform prefetching.
+ *
+ * The list can be iterated with&drm_gpuva_for_each_opand must be processed
+ * in the given order. It can contain prefetch operations.
+ *
+ * There can be an arbitrary amount of prefetch operations.
+ *
+ * After the caller finished processing the returned&drm_gpuva_ops, they must
+ * be freed with&drm_gpuva_ops_free.
+ *
+ * Returns: a pointer to the&drm_gpuva_opson success, anERR_PTRon failure
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +38
-0+struct drm_gpuva_ops * +drm_gpuva_prefetch_ops_create (struct drm_gpuva_manager *mgr, + u64 addr, u64 range) +{ + struct drm_gpuva_ops *ops; + struct drm_gpuva_op *op; + struct drm_gpuva *va; + u64 end = addr + range; + int ret; + + ops = kzalloc(sizeof(*ops), GFP_KERNEL); + if (!ops) + return ERR_PTR(-ENOMEM); + + INIT_LIST_HEAD(&ops->list); + + drm_gpuva_for_each_va_range(va, mgr, addr, end) { + op = gpuva_op_alloc(mgr); + if (!op) { + ret = -ENOMEM; + goto err_free_ops; + } + + op->op = DRM_GPUVA_OP_PREFETCH; + op->prefetch.va = va; + list_add_tail(&op->entry, &ops->list); + } + + return ops; + +err_free_ops: + drm_gpuva_ops_free(mgr, ops); + return ERR_PTR(ret); +} +EXPORT_SYMBOL (drm_gpuva_prefetch_ops_create); + +/** +* drm_gpuva_gem_unmap_ops_create - creates the &drm_gpuva_ops to unmap a GEM
+ * @mgr: the&drm_gpuva_managerrepresenting the GPU VA space
+ * @obj: the&drm_gem_objectto unmap
+ *
+ * This function creates a list of operations to perform unmapping for every
+ * GPUVA attached to a GEM.
+ *
+ * The list can be iterated with&drm_gpuva_for_each_opand consists out of an
+ * arbitrary amount of unmap operations.
+ *
+ * After the caller finished processing the returned&drm_gpuva_ops, they must
+ * be freed with&drm_gpuva_ops_free.
+ *
+ * It is the callers responsibility to protect the GEMs GPUVA list against
+ * concurrent access using the GEMsdma_resvlock.
+ *
+ * Returns: a pointer to the&drm_gpuva_opson success, anERR_PTRon failure
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +47
-0+struct drm_gpuva_ops * +drm_gpuva_gem_unmap_ops_create (struct drm_gpuva_manager *mgr, + struct drm_gem_object *obj) +{ + struct drm_gpuva_ops *ops; + struct drm_gpuva_op *op; + struct drm_gpuva *va; + int ret; + + if (drm_gpuva_manager_external_lock(mgr)) + drm_gpuva_manager_ext_assert_held(mgr); + else + dma_resv_assert_held(obj->resv); + + ops = kzalloc(sizeof(*ops), GFP_KERNEL); + if (!ops) + return ERR_PTR(-ENOMEM); + + INIT_LIST_HEAD(&ops->list); + + drm_gem_for_each_gpuva(va, obj) { + op = gpuva_op_alloc(mgr); + if (!op) { + ret = -ENOMEM; + goto err_free_ops; + } + + op->op = DRM_GPUVA_OP_UNMAP; + op->unmap.va = va; + list_add_tail(&op->entry, &ops->list); + } + + return ops; + +err_free_ops: + drm_gpuva_ops_free(mgr, ops); + return ERR_PTR(ret); +} +EXPORT_SYMBOL (drm_gpuva_gem_unmap_ops_create); + + +/** +* drm_gpuva_ops_free - free the given &drm_gpuva_ops +* @mgr: the &drm_gpuva_manager the ops were created for +* @ops: the &drm_gpuva_ops to free +* +* Frees the given &drm_gpuva_ops structure including all the ops associated
+ * with it.
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +21
-0+void +drm_gpuva_ops_free (struct drm_gpuva_manager *mgr, + struct drm_gpuva_ops *ops) +{ + struct drm_gpuva_op *op, *next; + + drm_gpuva_for_each_op_safe(op, next, ops) { + list_del(&op->entry); + + if (op->op == DRM_GPUVA_OP_REMAP) { + kfree(op->remap.prev); + kfree(op->remap.next); + kfree(op->remap.unmap); + } + + gpuva_op_free(mgr, op); + } + + kfree(ops); +} +EXPORT_SYMBOL (drm_gpuva_ops_free);include/drm/drm_drv.h +6
-0@@ -104,6 +104,12 @@ enum drm_driver_feature {* acceleration should be handled by two drivers that are connected using auxiliary bus. */ DRIVER_COMPUTE_ACCEL = BIT(7), + /** + * @DRIVER_GEM_GPUVA: + * + * Driver supports user defined GPU VA bindings for GEM objects. + */ + DRIVER_GEM_GPUVA = BIT(8), /* IMPORTANT: Below are all the legacy flags, add new ones above. */include/drm/drm_gem.h +21
-0@@ -36,6 +36,8 @@ #include <linux/kref.h> #include <linux/dma-resv.h> +#include <linux/list.h> +#include <linux/mutex.h> #include <drm/drm_vma_manager.h>@@ -379,6 +381,18 @@ struct drm_gem_object {*/ struct dma_resv _resv; + /** + * @gpuva: + * + * Provides the list of GPU VAs attached to this GEM object. + * + * Drivers should lock list accesses with the GEMs &dma_resv lock + * (&drm_gem_object.resv). + */ + struct { + struct list_head list; + } gpuva; + /*** @funcs: * @@ -526,4 +540,42 @@ unsigned long drm_gem_lru_scan(struct drm_gem_lru *lru, int drm_gem_evict(struct drm_gem_object *obj); +/** +* drm_gem_gpuva_init - initialize the gpuva list of a GEM object +* @obj: the &drm_gem_object +* +* This initializes the &drm_gem_object's &drm_gpuva list. +* +* Calling this function is only necessary for drivers intending to support the
+ *&drm_driver_featureDRIVER_GEM_GPUVA.
+ */include/drm/drm_gem.h unsigned long drm_gem_lru_scan(struct drm_gem_lru *lru, · line 558
Quoted 7 lines
include/drm/drm_gem.h +7
-0+static inline void drm_gem_gpuva_init (struct drm_gem_object *obj) +{ + INIT_LIST_HEAD(&obj->gpuva.list); +} + +/** +* drm_gem_for_each_gpuva - iternator to walk over a list of gpuvas s/iternator/iterator/. (multiple places) Although since "iterator" typically refers to an object being iterated over, perhaps
"
drm_gem_for_each_gpuva()- iterate over a list of gpuvas", (general comment across the patch).Quoted 851 lines
+ * @entry:
&drm_gpuvastructure to assign to in each iteration step
+ * @obj: the&drm_gem_objectthe&drm_gpuvasto walk are associated with
+ *
+ * This iterator walks over all&drm_gpuvastructures associated with the
+ *&drm_gpuva_manager.
+ */include/drm/drm_gem.h +5
-0+#define drm_gem_for_each_gpuva (entry__, obj__) \ + list_for_each_entry(entry__, &(obj__)->gpuva.list, gem.entry) + +/** +* drm_gem_for_each_gpuva_safe - iternator to safely walk over a list of gpuvas
+ * @entry:&drm_gpuvastructure to assign to in each iteration step
+ * @next: &next&drm_gpuvato store the next step
+ * @obj: the&drm_gem_objectthe&drm_gpuvasto walk are associated with
+ *
+ * This iterator walks over all&drm_gpuvastructures associated with the
+ *&drm_gem_object. It is implemented withlist_for_each_entry_safe(), hence
+ * it is save against removal of elements.
+ */
+#definedrm_gem_for_each_gpuva_safe(entry__, next__, obj__)\
+list_for_each_entry_safe(entry__, next__, &(obj__)->gpuva.list, gem.entry)
+
#endif /*__DRM_GEM_H__*/include/drm/drm_gpuva_mgr.h new file +9
-0@@ -0,0 +1,756 @@ +/* SPDX-License-Identifier: GPL-2.0 */ + +#ifndef __DRM_GPUVA_MGR_H__ +#define __DRM_GPUVA_MGR_H__ + +/* +* Copyright (c) 2022 Red Hat. +* +* Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */include/drm/drm_gpuva_mgr.h new file +84
-0+ +#include <linux/list.h> +#include <linux/rbtree.h> +#include <linux/types.h> + +#include <drm/drm_gem.h> + +struct drm_gpuva_manager ; +struct drm_gpuva_fn_ops ; + +/** +* enum drm_gpuva_flags - flags for struct drm_gpuva +*/ +enum drm_gpuva_flags { + /** + * @DRM_GPUVA_INVALIDATED: + * + * Flag indicating that the &drm_gpuva's backing GEM is invalidated. + */ + DRM_GPUVA_INVALIDATED = (1 << 0), + + /** + * @DRM_GPUVA_SPARSE: + * + * Flag indicating that the &drm_gpuva is a sparse mapping. + */ + DRM_GPUVA_SPARSE = (1 << 1), + + /** + * @DRM_GPUVA_USERBITS: user defined bits + */ + DRM_GPUVA_USERBITS = (1 << 2), +}; + +/** +* struct drm_gpuva - structure to track a GPU VA mapping +* +* This structure represents a GPU VA mapping and is associated with a +* &drm_gpuva_manager. +* +* Typically, this structure is embedded in bigger driver structures. +*/ +struct drm_gpuva { + /** + * @mgr: the &drm_gpuva_manager this object is associated with + */ + struct drm_gpuva_manager *mgr; + + /** + * @flags: the &drm_gpuva_flags for this mapping + */ + enum drm_gpuva_flags flags; + + /** + * @va: structure containing the address and range of the &drm_gpuva + */ + struct { + /** + * @addr: the start address + */ + u64 addr; + + /* + * @range: the range + */ + u64 range; + } va; + + /** + * @gem: structure containing the &drm_gem_object and it's offset + */ + struct { + /** + * @offset: the offset within the &drm_gem_object + */ + u64 offset; + + /** + * @obj: the mapped &drm_gem_object + */ + struct drm_gem_object *obj; + + /** + * @entry: the &list_head to attach this object to a
&drm_gem_object
+ */
+ structlist_headentry;
+ } gem;include/drm/drm_gpuva_mgr.h new file +21
-0+ + /** + * @rb: structure containing data to store &drm_gpuvas in a rb-tree + */ + struct { + /** + * @rb: the rb-tree node + */ + struct rb_node node; + + /** + * @entry: The &list_head to additionally connect &drm_gpuvas + * in the same order they appear in the interval tree. This is + * useful to keep iterating &drm_gpuvas from a start node found + * through the rb-tree while doing modifications on the rb-tree + * itself. + */ + struct list_head entry; + + /** + * @__subtree_last: needed by the interval tree, holdinglast-in-subtree
+ */
+ u64__subtree_last;
+ } rb;include/drm/drm_gpuva_mgr.h new file +3
-0+}; + +int drm_gpuva_insert(struct drm_gpuva_manager *mgr, struct drm_gpuva*va);
include/drm/drm_gpuva_mgr.h new file +10
-0+void drm_gpuva_remove (struct drm_gpuva *va); + +void drm_gpuva_link (struct drm_gpuva *va); +void drm_gpuva_unlink (struct drm_gpuva *va); + +struct drm_gpuva *drm_gpuva_find (struct drm_gpuva_manager *mgr, + u64 addr, u64 range); +struct drm_gpuva *drm_gpuva_find_first (struct drm_gpuva_manager *mgr, + u64 addr, u64 range); +struct drm_gpuva *drm_gpuva_find_prev(struct drm_gpuva_manager *mgr,u64 start);
+structdrm_gpuva*drm_gpuva_find_next(structdrm_gpuva_manager*mgr, u64 end);include/drm/drm_gpuva_mgr.h new file +2
-0+ +bool drm_gpuva_interval_empty(struct drm_gpuva_manager *mgr, u64addr, u64 range);
include/drm/drm_gpuva_mgr.h new file +2
-0+ +static inline void drm_gpuva_init(struct drm_gpuva *va, u64 addr, u64range,
+ structdrm_gem_object*obj, u64 offset)include/drm/drm_gpuva_mgr.h new file +9
-0+{ + va->va.addr = addr; + va->va.range = range; + va->gem.obj = obj; + va->gem.offset = offset; +} + +/** +* drm_gpuva_invalidate - sets whether the backing GEM of this
&drm_gpuvais
+ * invalidated
+ * @va: the&drm_gpuvato set the invalidate flag for
+ * @invalidate: indicates whether the&drm_gpuvais invalidated
+ */
+static inline voiddrm_gpuva_invalidate(struct drm_gpuva *va, bool invalidate)include/drm/drm_gpuva_mgr.h new file +9
-0+{ + if (invalidate) + va->flags |= DRM_GPUVA_INVALIDATED; + else + va->flags &= ~DRM_GPUVA_INVALIDATED; +} + +/** +* drm_gpuva_invalidated - indicates whether the backing BO of this
&drm_gpuva
+ * is invalidated
+ * @va: the&drm_gpuvato check
+ */include/drm/drm_gpuva_mgr.h new file +16
-0+static inline bool drm_gpuva_invalidated (struct drm_gpuva *va) +{ + return va->flags & DRM_GPUVA_INVALIDATED; +} + +#ifdef CONFIG_LOCKDEP +typedef struct lockdep_map *lockdep_map_p ; +#define drm_gpuva_manager_ext_assert_held (mgr) \ + lockdep_assert(lock_is_held((mgr)->ext_lock) != LOCK_STATE_NOT_HELD) +/** +* drm_gpuva_manager_set_ext_lock - set the external lock according to +* @DRM_GPUVA_MANAGER_LOCK_EXTERN +* @mgr: the &drm_gpuva_manager to set the lock for +* @lock: the lock to set +* +* If @DRM_GPUVA_MANAGER_LOCK_EXTERN is set, drivers need to call this function
+ * to provide the lock used to lock linking and unlinking of&drm_gpuvasto the
+ *&drm_gem_objectsGPUVA list.
+ */include/drm/drm_gpuva_mgr.h new file +5
-0+#define drm_gpuva_manager_set_ext_lock (mgr, lock) \ + (mgr)->ext_lock = &(lock)->dep_map +#else +typedef struct {/* nothing */ }lockdep_map_p ; +#define drm_gpuva_manager_ext_assert_held (mgr) do {(void)(mgr); } while (0)
include/drm/drm_gpuva_mgr.h new file +5
-0+#define drm_gpuva_manager_set_ext_lock (mgr, lock) do { } while (0) +#endif + +/** +* enum drm_gpuva_manager_flags - the feature flags for the
&drm_gpuva_manager
+ */include/drm/drm_gpuva_mgr.h new file +17
-0+enum drm_gpuva_manager_flags { + /** + * @DRM_GPUVA_MANAGER_LOCK_EXTERN: + * + * Indicates the driver has it's own external lock for linking and + * unlinking &drm_gpuvas to the &drm_gem_objects GPUVA list. + * + * When setting this flag it is rquired to set a lock via + * drm_gpuva_set_ext_lock(). + */ + DRM_GPUVA_MANAGER_LOCK_EXTERN = (1 << 0), +}; + +/** +* struct drm_gpuva_manager - DRM GPU VA Manager +* +* The DRM GPU VA Manager keeps track of a GPU's virtual address space by using
+ *&maple_treestructures. Typically, this structure is embedded in bigger
+ * driver structures.
+ *
+ * Drivers can pass addresses and ranges in an arbitrary unit, e.g. bytes or
+ * pages.
+ *
+ * There should be one manager instance per GPU virtual address space.
+ */include/drm/drm_gpuva_mgr.h new file +46
-0+struct drm_gpuva_manager { + /** + * @name: the name of the DRM GPU VA space + */ + const char *name; + + /** + * @flags: the feature flags of the &drm_gpuva_manager + */ + enum drm_gpuva_manager_flags flags; + + /** + * @mm_start: start of the VA space + */ + u64 mm_start; + + /** + * @mm_range: length of the VA space + */ + u64 mm_range; + + /** + * @rb: structures to track &drm_gpuva entries + */ + struct { + /** + * @tree: the rb-tree to track GPU VA mappings + */ + struct rb_root_cached tree; + + /** + * @list: the &list_head to track GPU VA mappings + */ + struct list_head list; + } rb; + + /** + * @kernel_alloc_node: + * + * &drm_gpuva representing the address space cutout reserved for + * the kernel + */ + struct drm_gpuva kernel_alloc_node; + + /** + * @ops: &drm_gpuva_fn_ops providing the split/merge steps todrivers
+ */
+ const structdrm_gpuva_fn_ops*ops;
+
+ /**
+ * @ext_lock:&lockdep_mapaccording to
@DRM_GPUVA_MANAGER_LOCK_EXTERN
+ */
+lockdep_map_pext_lock;include/drm/drm_gpuva_mgr.h new file +16
-0+}; + +void drm_gpuva_manager_init (struct drm_gpuva_manager *mgr, + const char *name, + u64 start_offset, u64 range, + u64 reserve_offset, u64 reserve_range, + const struct drm_gpuva_fn_ops *ops, + enum drm_gpuva_manager_flags flags); +void drm_gpuva_manager_destroy (struct drm_gpuva_manager *mgr); + +/** +* drm_gpuva_manager_external_lock - indicates whether the +* @DRM_GPUVA_MANAGER_LOCK_EXTERN flag is set +* @mgr: the &drm_gpuva_manager to check the flag for +*/ +static inline bool drm_gpuva_manager_external_lock(struct
drm_gpuva_manager*mgr)include/drm/drm_gpuva_mgr.h new file +6
-0+{ + return mgr->flags & DRM_GPUVA_MANAGER_LOCK_EXTERN; +} + +/** +* drm_gpuva_for_each_va_range - iternator to walk over a range of
&drm_gpuvas
+ * @va__:&drm_gpuvastructure to assign to in each iteration step
+ * @mgr__:&drm_gpuva_managerto walk over
+ * @start__: starting offset, the first gpuva will overlap this
+ * @end__: ending offset, the last gpuva will start before this (but may
+ * overlap)
+ *
+ * This iterator walks over all&drm_gpuvasin the&drm_gpuva_managerthat lie
+ * between @start__ and @end__. It is implemented similarly tolist_for_each(),
+ * but is using the &drm_gpuva_manager's internal interval tree to accelerate
+ * the search for the starting&drm_gpuva, and hence isn't safe against removal
+ * of elements. It assumes that @end__ is within (or is the upper limit of) the
+ *&drm_gpuva_manager. This iterator does not skip over the &drm_gpuva_manager's
+ * @kernel_alloc_node.
+ */include/drm/drm_gpuva_mgr.h new file +8
-0+#define drm_gpuva_for_each_va_range (va__, mgr__, start__, end__) \ + for (va__ = drm_gpuva_find_first((mgr__), (start__), (end__)); \ + va__ && (va__->va.addr < (end__)) && \ + !list_entry_is_head(va__, &(mgr__)->rb.list, rb.entry); \ + va__ = list_next_entry(va__, rb.entry)) + +/** +* drm_gpuva_for_each_va_range_safe - iternator to safely walk over a range of
+ *&drm_gpuvas
+ * @va__:&drm_gpuvato assign to in each iteration step
+ * @next__: another&drm_gpuvato use as temporary storage
+ * @mgr__:&drm_gpuva_managerto walk over
+ * @start__: starting offset, the first gpuva will overlap this
+ * @end__: ending offset, the last gpuva will start before this (but may
+ * overlap)
+ *
+ * This iterator walks over all&drm_gpuvasin the&drm_gpuva_managerthat lie
+ * between @start__ and @end__. It is implemented similarly to
+ *list_for_each_safe(), but is using the &drm_gpuva_manager's internal interval
+ * tree to accelerate the search for the starting&drm_gpuva, and hence is safe
+ * against removal of elements. It assumes that @end__ is within (or is the
+ * upper limit of) the&drm_gpuva_manager. This iterator does not skip over the
+ * &drm_gpuva_manager's @kernel_alloc_node.
+ */
+#definedrm_gpuva_for_each_va_range_safe(va__, next__, mgr__, start__, end__)\
+ for (va__=drm_gpuva_find_first((mgr__), (start__), (end__)), \
+next__= va ?list_next_entry(va__, rb.entry): NULL; \
+va__&& (va__->va.addr< (end__)) && \
+ !list_entry_is_head(va__, &(mgr__)->rb.list, rb.entry); \
+va__=next__,next__=list_next_entry(va__, rb.entry))include/drm/drm_gpuva_mgr.h new file +7
-0+ +/** +* drm_gpuva_for_each_va - iternator to walk over all &drm_gpuvas +* @va__: &drm_gpuva to assign to in each iteration step +* @mgr__: &drm_gpuva_manager to walk over +* +* This iterator walks over all &drm_gpuva structures associated with the given
+ *&drm_gpuva_manager.
+ */include/drm/drm_gpuva_mgr.h new file +5
-0+#define drm_gpuva_for_each_va (va__, mgr__) \ + list_for_each_entry(va__, &(mgr__)->rb.list, rb.entry) + +/** +* drm_gpuva_for_each_va_safe - iternator to safely walk over all
&drm_gpuvas
+ * @va__:&drm_gpuvato assign to in each iteration step
+ * @next__: another&drm_gpuvato use as temporary storage
+ * @mgr__:&drm_gpuva_managerto walk over
+ *
+ * This iterator walks over all&drm_gpuvastructures associated with the given
+ *&drm_gpuva_manager. It is implemented withlist_for_each_entry_safe(), and
+ * hence safe against the removal of elements.
+ */include/drm/drm_gpuva_mgr.h new file +7
-0+#define drm_gpuva_for_each_va_safe (va__, next__, mgr__) \ + list_for_each_entry_safe(va__, next__, &(mgr__)->rb.list, rb.entry) + +/** +* enum drm_gpuva_op_type - GPU VA operation type +* +* Operations to alter the GPU VA mappings tracked by the
&drm_gpuva_manager.
+ */include/drm/drm_gpuva_mgr.h new file +77
-0+enum drm_gpuva_op_type { + /** + * @DRM_GPUVA_OP_MAP: the map op type + */ + DRM_GPUVA_OP_MAP, + + /** + * @DRM_GPUVA_OP_REMAP: the remap op type + */ + DRM_GPUVA_OP_REMAP, + + /** + * @DRM_GPUVA_OP_UNMAP: the unmap op type + */ + DRM_GPUVA_OP_UNMAP, + + /** + * @DRM_GPUVA_OP_PREFETCH: the prefetch op type + */ + DRM_GPUVA_OP_PREFETCH, +}; + +/** +* struct drm_gpuva_op_map - GPU VA map operation +* +* This structure represents a single map operation generated by the +* DRM GPU VA manager. +*/ +struct drm_gpuva_op_map { + /** + * @va: structure containing address and range of a map + * operation + */ + struct { + /** + * @addr: the base address of the new mapping + */ + u64 addr; + + /** + * @range: the range of the new mapping + */ + u64 range; + } va; + + /** + * @gem: structure containing the &drm_gem_object and it's offset + */ + struct { + /** + * @offset: the offset within the &drm_gem_object + */ + u64 offset; + + /** + * @obj: the &drm_gem_object to map + */ + struct drm_gem_object *obj; + } gem; +}; + +/** +* struct drm_gpuva_op_unmap - GPU VA unmap operation +* +* This structure represents a single unmap operation generated by the +* DRM GPU VA manager. +*/ +struct drm_gpuva_op_unmap { + /** + * @va: the &drm_gpuva to unmap + */ + struct drm_gpuva *va; + + /** +* @keep: +* +* Indicates whether this &drm_gpuva is physically contiguous with the
+ * original mapping request.
+ *
+ * Optionally, if &keep is set, drivers may keep the actual page table
+ * mappings for this&drm_gpuva, adding the missing page table entries
+ * only and update the&drm_gpuva_manageraccordingly. + */
+ bool keep;include/drm/drm_gpuva_mgr.h new file +6
-0+}; + +/** +* struct drm_gpuva_op_remap - GPU VA remap operation +* +* This represents a single remap operation generated by the DRM GPU VA manager.
+ *
+ * A remap operation is generated when an existing GPU VA mmapping is split up
+ * by inserting a new GPU VA mapping or by partially unmapping existent
+ * mapping(s), hence it consists of a maximum of two map and one unmap
+ * operation.
+ *
+ * The @unmap operation takes care of removing the original existing mapping.
+ * @prev is used to remap the preceding part, @next the subsequent part.
+ *
+ * If either a new mapping's start address is aligned with the start address
+ * of the old mapping or the new mapping's end address is aligned with the
+ * end address of the old mapping, either @prev or @next is NULL.
+ *
+ * Note, the reason for a dedicated remap operation, rather than arbitrary
+ * unmap and map operations, is to give drivers the chance of extracting driver
+ * specific data for creating the new mappings from the unmap operations's
+ *&drm_gpuvastructure which typically is embedded in larger driver specific
+ * structures.
+ */include/drm/drm_gpuva_mgr.h new file +21
-0+struct drm_gpuva_op_remap { + /** + * @prev: the preceding part of a split mapping + */ + struct drm_gpuva_op_map *prev; + + /** + * @next: the subsequent part of a split mapping + */ + struct drm_gpuva_op_map *next; + + /** + * @unmap: the unmap operation for the original existing mapping + */ + struct drm_gpuva_op_unmap *unmap; +}; + +/** +* struct drm_gpuva_op_prefetch - GPU VA prefetch operation +* +* This structure represents a single prefetch operation generated by the
+ * DRM GPU VA manager.
+ */include/drm/drm_gpuva_mgr.h new file +69
-0+struct drm_gpuva_op_prefetch { + /** + * @va: the &drm_gpuva to prefetch + */ + struct drm_gpuva *va; +}; + +/** +* struct drm_gpuva_op - GPU VA operation +* +* This structure represents a single generic operation. +* +* The particular type of the operation is defined by @op. +*/ +struct drm_gpuva_op { + /** + * @entry: + * + * The &list_head used to distribute instances of this struct within + * &drm_gpuva_ops. + */ + struct list_head entry; + + /** + * @op: the type of the operation + */ + enum drm_gpuva_op_type op; + + union { + /** + * @map: the map operation + */ + struct drm_gpuva_op_map map; + + /** + * @remap: the remap operation + */ + struct drm_gpuva_op_remap remap; + + /** + * @unmap: the unmap operation + */ + struct drm_gpuva_op_unmap unmap; + + /** + * @prefetch: the prefetch operation + */ + struct drm_gpuva_op_prefetch prefetch; + }; +}; + +/** +* struct drm_gpuva_ops - wraps a list of &drm_gpuva_op +*/ +struct drm_gpuva_ops { + /** + * @list: the &list_head + */ + struct list_head list; +}; + +/** +* drm_gpuva_for_each_op - iterator to walk over &drm_gpuva_ops +* @op: &drm_gpuva_op to assign in each iteration step +* @ops: &drm_gpuva_ops to walk +* +* This iterator walks over all ops within a given list of operations. +*/ +#define drm_gpuva_for_each_op (op, ops) list_for_each_entry(op,&(ops)->list, entry)
include/drm/drm_gpuva_mgr.h new file +3
-0+ +/** +* drm_gpuva_for_each_op_safe - iterator to safely walk over
&drm_gpuva_ops
+ * @op:&drm_gpuva_opto assign in each iteration step
+ * @next: &next&drm_gpuva_opto store the next step
+ * @ops:&drm_gpuva_opsto walk
+ *
+ * This iterator walks over all ops within a given list of operations. It is
+ * implemented withlist_for_each_safe(), so save against removal of elements.
+ */include/drm/drm_gpuva_mgr.h new file +5
-0+#define drm_gpuva_for_each_op_safe (op, next, ops) \ + list_for_each_entry_safe(op, next, &(ops)->list, entry) + +/** +* drm_gpuva_for_each_op_from_reverse - iterate backwards from the given point
+ * @op:&drm_gpuva_opto assign in each iteration step
+ * @ops:&drm_gpuva_opsto walk
+ *
+ * This iterator walks over all ops within a given list of operations beginning
+ * from the given operation in reverse order.
+ */include/drm/drm_gpuva_mgr.h new file +5
-0+#define drm_gpuva_for_each_op_from_reverse (op, ops) \ + list_for_each_entry_from_reverse(op, &(ops)->list, entry) + +/** +* drm_gpuva_first_op - returns the first &drm_gpuva_op from
&drm_gpuva_ops
+ * @ops: the&drm_gpuva_opsto get the fist&drm_gpuva_opfrom
+ */include/drm/drm_gpuva_mgr.h new file +5
-0+#define drm_gpuva_first_op (ops) \ + list_first_entry(&(ops)->list, struct drm_gpuva_op, entry) + +/** +* drm_gpuva_last_op - returns the last &drm_gpuva_op from
&drm_gpuva_ops
+ * @ops: the&drm_gpuva_opsto get the last&drm_gpuva_opfrom
+ */include/drm/drm_gpuva_mgr.h new file +46
-0+#define drm_gpuva_last_op (ops) \ + list_last_entry(&(ops)->list, struct drm_gpuva_op, entry) + +/** +* drm_gpuva_prev_op - previous &drm_gpuva_op in the list +* @op: the current &drm_gpuva_op +*/ +#define drm_gpuva_prev_op (op) list_prev_entry(op, entry) + +/** +* drm_gpuva_next_op - next &drm_gpuva_op in the list +* @op: the current &drm_gpuva_op +*/ +#define drm_gpuva_next_op (op) list_next_entry(op, entry) + +struct drm_gpuva_ops * +drm_gpuva_sm_map_ops_create (struct drm_gpuva_manager *mgr, + u64 addr, u64 range, + struct drm_gem_object *obj, u64 offset); +struct drm_gpuva_ops * +drm_gpuva_sm_unmap_ops_create (struct drm_gpuva_manager *mgr, + u64 addr, u64 range); + +struct drm_gpuva_ops * +drm_gpuva_prefetch_ops_create (struct drm_gpuva_manager *mgr, + u64 addr, u64 range); + +struct drm_gpuva_ops * +drm_gpuva_gem_unmap_ops_create (struct drm_gpuva_manager *mgr, + struct drm_gem_object *obj); + +void drm_gpuva_ops_free (struct drm_gpuva_manager *mgr, + struct drm_gpuva_ops *ops); + +static inline void drm_gpuva_init_from_op (struct drm_gpuva *va, + struct drm_gpuva_op_map *op) +{ + drm_gpuva_init(va, op->va.addr, op->va.range, + op->gem.obj, op->gem.offset); +} + +/** +* struct drm_gpuva_fn_ops - callbacks for split/merge steps +* +* This structure defines the callbacks used by &drm_gpuva_sm_map and +* &drm_gpuva_sm_unmap to provide the split/merge steps for map and unmap
+ * operations to drivers.
+ */include/drm/drm_gpuva_mgr.h new file +41
-0+struct drm_gpuva_fn_ops { + /** + * @op_alloc: called when the &drm_gpuva_manager allocates + * a struct drm_gpuva_op + * + * Some drivers may want to embed struct drm_gpuva_op into driver + * specific structures. By implementing this callback drivers can + * allocate memory accordingly. + * + * This callback is optional. + */ + struct drm_gpuva_op *(*op_alloc)(void); + + /** + * @op_free: called when the &drm_gpuva_manager frees a + * struct drm_gpuva_op + * + * Some drivers may want to embed struct drm_gpuva_op into driver + * specific structures. By implementing this callback drivers can + * free the previously allocated memory accordingly. + * + * This callback is optional. + */ + void (*op_free)(struct drm_gpuva_op *op); + + /** + * @sm_step_map: called from &drm_gpuva_sm_map to finally insert the + * mapping once all previous steps were completed + * + * The &priv pointer matches the one the driver passed to + * &drm_gpuva_sm_map or &drm_gpuva_sm_unmap, respectively. + * + * Can be NULL if &drm_gpuva_sm_map is used. + */ + int (*sm_step_map)(struct drm_gpuva_op *op, void *priv); + + /** + * @sm_step_remap: called from &drm_gpuva_sm_map and + * &drm_gpuva_sm_unmap to split up an existent mapping + * + * This callback is called when existent mapping needs to besplit up.
+ * This is the case when either a newly requested mapping overlaps or
+ * is enclosed by an existent mapping or a partial unmap of an existent
+ * mapping is requested.
+ *
+ * The &priv pointer matches the one the driver passed to + *&drm_gpuva_sm_mapor&drm_gpuva_sm_unmap, respectively. + *
+ * Can be NULL if neither&drm_gpuva_sm_mapnor&drm_gpuva_sm_unmapis
+ * used.
+ */
+ int (*sm_step_remap)(structdrm_gpuva_op*op, void *priv); +
+ /**
+ * @sm_step_unmap: called from&drm_gpuva_sm_mapand
+ *&drm_gpuva_sm_unmapto unmap an existent mapping
+ *
+ * This callback is called when existent mapping needs to be unmapped.
+ * This is the case when either a newly requested mapping encloses an
+ * existent mapping or an unmap of an existent mapping is requested. + *
+ * The &priv pointer matches the one the driver passed to + *&drm_gpuva_sm_mapor&drm_gpuva_sm_unmap, respectively. + *
+ * Can be NULL if neither&drm_gpuva_sm_mapnor&drm_gpuva_sm_unmapis
+ * used.
+ */
+ int (*sm_step_unmap)(structdrm_gpuva_op*op, void *priv);include/drm/drm_gpuva_mgr.h new file +2
-0+}; +include/drm/drm_gpuva_mgr.h line 750
Quoted 10 lines
include/drm/drm_gpuva_mgr.h new file +10
-0+int drm_gpuva_sm_map (struct drm_gpuva_manager *mgr,void *priv, + u64 addr, u64 range, + struct drm_gem_object *obj, u64 offset); + +int drm_gpuva_sm_unmap (struct drm_gpuva_manager *mgr,void *priv, + u64 addr, u64 range); + +void drm_gpuva_map (struct drm_gpuva_manager *mgr, + struct drm_gpuva *va, + struct drm_gpuva_op_map *op);Missing newline
Quoted 3 lines
+void drm_gpuva_remap(struct
drm_gpuva*prev,
+ structdrm_gpuva*next,
+ structdrm_gpuva_op_remap*op);Missing newline
Quoted 3 lines
include/drm/drm_gpuva_mgr.h new file +3
-0+void drm_gpuva_unmap (struct drm_gpuva_op_unmap *op); + +#endif /* __DRM_GPUVA_MGR_H__ */
Re: [PATCH drm-next v6 02/13] drm: manager to keep track of GPUs VA mappings
Quoted 638 lines · Danilo Krummrich
Hi Thomas,
Quoted 626 lines · Thomas Hellström (Intel)
Hi, Danilo
Some review comments below:
Quoted 10 lines · Danilo Krummrich
Add infrastructure to keep track of GPU virtual address (VA) mappings with a decicated VA space manager implementation.
New UAPIs, motivated by Vulkan sparse memory bindings graphics drivers start implementing, allow userspace applications to request multiple and arbitrary GPU VA mappings of buffer objects. The DRM GPU VA manager is intended to serve the following purposes in this context.
1) Provide infrastructure to track GPU VA allocations and mappings, making use of the
maple_tree.It looks like we're not using the maple tree anymore, but rather an instantiation of an interval tree.
(Perhaps as a follow-up it makes sense to provide a pre-instantiated common u64 version of the interval tree in addition to the unsigned long one since it appears to be used in multiple places in graphics drivers).
Quoted 102 lines
2) Generically connect GPU VA mappings to their backing buffers, in particular DRM GEM objects.
3) Provide a common implementation to perform more complex mapping operations on the GPU VA space. In particular splitting and merging of GPU VA mappings, e.g. for intersecting mapping requests or partial unmap requests.
Tested-by Donald Robson<[email protected]>Reviewed-by Boris Brezillon<[email protected]>Suggested-by Dave Airlie<[email protected]>Signed-off-by Danilo Krummrich<[email protected]>---
Documentation/gpu/drm-mm.rst | 36 + drivers/gpu/drm/Makefile | 1 + drivers/gpu/drm/drm_gem.c | 3 +drivers/gpu/drm/drm_gpuva_mgr.c| 1743
+++++++++++++++++++++++++++++++
include/drm/drm_drv.h | 6 + include/drm/drm_gem.h | 52 + include/drm/drm_gpuva_mgr.h | 756 ++++++++++++++7 files changed, 2597 insertions(+)create mode 100644
drivers/gpu/drm/drm_gpuva_mgr.ccreate mode 100644include/drm/drm_gpuva_mgr.hDocumentation/gpu/drm-mm.rst +0
-0@@ -466,6 +466,42 @@ DRM MM Range Allocator Function References.. kernel-doc::
drivers/gpu/drm/drm_mm.c
:export:Documentation/gpu/drm-mm.rst +36
-0+DRM GPU VA Manager +================== + +Overview +-------- + +.. kernel-doc:: drivers/gpu/drm/drm_gpuva_mgr.c + :doc: Overview + +Split and Merge +--------------- + +.. kernel-doc:: drivers/gpu/drm/drm_gpuva_mgr.c + :doc: Split and Merge + +Locking +------- + +.. kernel-doc:: drivers/gpu/drm/drm_gpuva_mgr.c + :doc: Locking + +Examples +-------- + +.. kernel-doc:: drivers/gpu/drm/drm_gpuva_mgr.c + :doc: Examples + +DRM GPU VA Manager Function References +-------------------------------------- + +.. kernel-doc:: include/drm/drm_gpuva_mgr.h + :internal: + +.. kernel-doc:: drivers/gpu/drm/drm_gpuva_mgr.c + :export: +DRM Buddy Allocator
===================drivers/gpu/drm/Makefile +1
-0@@ -45,6 +45,7 @@ drm-y := \ drm_vblank.o \ drm_vblank_work.o \ drm_vma_manager.o \ + drm_gpuva_mgr.o \drm_writeback.o
drm-$(CONFIG_DRM_LEGACY) += \
drm_agpsupport.o \drivers/gpu/drm/drm_gem.c +3
-0@@ -164,6 +164,9 @@ void drm_gem_private_object_init(struct drm_device *dev, if (!obj->resv) obj->resv = &obj->_resv; + if (drm_core_check_feature(dev, DRIVER_GEM_GPUVA)) + drm_gem_gpuva_init(obj); + drm_vma_node_reset(&obj->vma_node); INIT_LIST_HEAD(&obj->lru_node); }drivers/gpu/drm/drm_gpuva_mgr.c new file +0
-0drivers/gpu/drm/drm_gpuva_mgr.c line 2
Quoted 2 lines
drivers/gpu/drm/drm_gpuva_mgr.c new file +2
-0@@ -0,0 +1,1743 @@ +// SPDX-License-Identifier: GPL-2.0 +/* SPDX-License is GPL-2.0 but below looks like MIT. Can we use "GPL-2.0 OR MIT" or does something restrict it to GPL-only?
Quoted 73 lines
drivers/gpu/drm/drm_gpuva_mgr.c new file +3
-0+* Copyright (c) 2022 Red Hat. +* +* Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors:
+ * Danilo Krummrich<[email protected]>
+ *
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +10
-0+ +#include <drm/drm_gpuva_mgr.h> + +#include <linux/interval_tree_generic.h> +#include <linux/mm.h> + +/** +* DOC: Overview +* +* The DRM GPU VA Manager, represented by struct drm_gpuva_manager keeps track
+ * of a GPU's virtual address (VA) space and manages the corresponding virtual
+ * mappings represented by&drm_gpuvaobjects. It also keeps track of the
+ * mapping's backing&drm_gem_objectbuffers.
+ *
+ *&drm_gem_objectbuffers maintain a list of&drm_gpuvaobjects representing
+ * all existent GPU VA mappings using this&drm_gem_objectas backing buffer.
+ *
+ * GPU VAs can be flagged as sparse, such that drivers may use GPU VAs to also
+ * keep track of sparse PTEs in order to support Vulkan 'Sparse Resources'.
+ *
+ * The GPU VA manager internally uses a rb-tree to manage the
+ *&drm_gpuvamappings within a GPU's virtual address space.
+ *
+ * The&drm_gpuva_managercontains a special&drm_gpuvarepresenting the
+ * portion of VA space reserved by the kernel. This node is initialized together
+ * with the GPU VA manager instance and removed when the GPU VA manager isQuoted 8 lines
+ * destroyed.
+ *
+ * In a typical application drivers would embed structdrm_gpuva_managerand
+ * structdrm_gpuvawithin their own driver specific structures, there won't be
+ * any memory allocations of it's own nor memory allocations of&drm_gpuvas/it's/its/
Quoted 9 lines
+ * entries.
+ *
+ * The data structures needed to store&drm_gpuvaswithin the&drm_gpuva_manager
+ * are contained within structdrm_gpuvaalready. Hence, for inserting
+ *&drm_gpuvaentries from within dma-fence signalling critical sections it is
+ * enough to pre-allocate the&drm_gpuvastructures.
+ */drivers/gpu/drm/drm_gpuva_mgr.c line 69
Quoted 5 lines
drivers/gpu/drm/drm_gpuva_mgr.c new file +5
-0+ +/** +* DOC: Split and Merge +* +* Besides it's capability to manage and represent a GPU VA space, the s/it's/its/
Quoted 178 lines
+ *
&drm_gpuva_manageralso provides functions to let the&drm_gpuva_manager
+ * calculate a sequence of operations to satisfy a given map or unmap request.
+ *
+ * Therefore the DRM GPU VA manager provides an algorithm implementing splitting
+ * and merging of existent GPU VA mappings with the ones that are requested to
+ * be mapped or unmapped. This feature is required by the Vulkan API to
+ * implement Vulkan 'Sparse Memory Bindings' - drivers UAPIs often refer to this
+ * as VM BIND.
+ *
+ * Drivers can calldrm_gpuva_sm_map()to receive a sequence of callbacks
+ * containing map, unmap and remap operations for a given newly requested
+ * mapping. The sequence of callbacks represents the set of operations to
+ * execute in order to integrate the new mapping cleanly into the current state
+ * of the GPU VA space.
+ *
+ * Depending on how the new GPU VA mapping intersects with the existent mappings
+ * of the GPU VA space the&drm_gpuva_fn_opscallbacks contain an arbitrary
+ * amount of unmap operations, a maximum of two remap operations and a single
+ * map operation. The caller might receive no callback at all if no operation is
+ * required, e.g. if the requested mapping already exists in the exact same way.
+ *
+ * The single map operation represents the original map operation requested by
+ * the caller.
+ *
+ *&drm_gpuva_op_unmapcontains a 'keep' field, which indicates whether the
+ *&drm_gpuvato unmap is physically contiguous with the original mapping
+ * request. Optionally, if 'keep' is set, drivers may keep the actual page table
+ * entries for this&drm_gpuva, adding the missing page table entries only and
+ * update the &drm_gpuva_manager's view of things accordingly.
+ *
+ * Drivers may do the same optimization, namely delta page table updates, also
+ * for remap operations. This is possible since&drm_gpuva_op_remapconsists of
+ * one unmap operation and one or two map operations, such that drivers can
+ * derive the page table update delta accordingly.
+ *
+ * Note that there can't be more than two existent mappings to split up, one at
+ * the beginning and one at the end of the new mapping, hence there is a
+ * maximum of two remap operations.
+ *
+ * Analogous todrm_gpuva_sm_map()drm_gpuva_sm_unmap()uses&drm_gpuva_fn_ops
+ * to call back into the driver in order to unmap a range of GPU VA space. The
+ * logic behind this function is way simpler though: For all existent mappings
+ * enclosed by the given range unmap operations are created. For mappings which
+ * are only partically located within the given range, remap operations are
+ * created such that those mappings are split up and re-mapped partically.
+ *
+ * As an alternative todrm_gpuva_sm_map()anddrm_gpuva_sm_unmap(),
+ *drm_gpuva_sm_map_ops_create()anddrm_gpuva_sm_unmap_ops_create()can be used
+ * to directly obtain an instance of structdrm_gpuva_opscontaining a list of
+ *&drm_gpuva_op, which can be iterated withdrm_gpuva_for_each_op(). This list
+ * contains the&drm_gpuva_opsanalogous to the callbacks one would receive when
+ * callingdrm_gpuva_sm_map()ordrm_gpuva_sm_unmap(). While this way requires
+ * more memory (to allocate the&drm_gpuva_ops), it provides drivers a way to
+ * iterate the&drm_gpuva_opmultiple times, e.g. once in a context where memory
+ * allocations are possible (e.g. to allocate GPU page tables) and once in the
+ * dma-fence signalling critical path.
+ *
+ * To update the &drm_gpuva_manager's view of the GPU VA space
+ *drm_gpuva_insert()anddrm_gpuva_remove()may be used. These functions can
+ * safely be used from&drm_gpuva_fn_opscallbacks originating from
+ *drm_gpuva_sm_map()ordrm_gpuva_sm_unmap(). However, it might be more
+ * convenient to use the provided helper functionsdrm_gpuva_map(),
+ *drm_gpuva_remap()anddrm_gpuva_unmap()instead.
+ *
+ * The following diagram depicts the basic relationships of existent GPU VA
+ * mappings, a newly requested mapping and the resulting mappings as implemented
+ * bydrm_gpuva_sm_map()- it doesn't cover any arbitrary combinations of these.
+ *
+ * 1) Requested mapping is identical. Replace it, but indicate the backing PTEs
+ * could be kept.
+ *
+ * ::
+ *
+ * 0 a 1
+ * old: |-----------| (bo_offset=n)
+ *
+ * 0 a 1
+ * req: |-----------| (bo_offset=n)
+ *
+ * 0 a 1
+ * new: |-----------| (bo_offset=n)
+ *
+ *
+ * 2) Requested mapping is identical, except for the BO offset, hence replace
+ * the mapping.
+ *
+ * ::
+ *
+ * 0 a 1
+ * old: |-----------| (bo_offset=n)
+ *
+ * 0 a 1
+ * req: |-----------| (bo_offset=m)
+ *
+ * 0 a 1
+ * new: |-----------| (bo_offset=m)
+ *
+ *
+ * 3) Requested mapping is identical, except for the backing BO, hence replace
+ * the mapping.
+ *
+ * ::
+ *
+ * 0 a 1
+ * old: |-----------| (bo_offset=n)
+ *
+ * 0 b 1
+ * req: |-----------| (bo_offset=n)
+ *
+ * 0 b 1
+ * new: |-----------| (bo_offset=n)
+ *
+ *
+ * 4) Existent mapping is a left aligned subset of the requested one, hence
+ * replace the existent one.
+ *
+ * ::
+ *
+ * 0 a 1
+ * old: |-----| (bo_offset=n)
+ *
+ * 0 a 2
+ * req: |-----------| (bo_offset=n)
+ *
+ * 0 a 2
+ * new: |-----------| (bo_offset=n)
+ *
+ * .. note::
+ * We expect to see the same result for a request with a different BOQuoted 8 lines
+ * and/or non-contiguous BO offset.
+ *
+ *
+ * 5) Requested mapping's range is a left aligned subset of the existent one,
+ * but backed by a different BO. Hence, map the requested mapping and split
+ * the existent one adjusting it's BO offset.Typo: s/it's/its/ above and in multiple places below.
Quoted 217 lines
+ *
+ * ::
+ *
+ * 0 a 2
+ * old: |-----------| (bo_offset=n)
+ *
+ * 0 b 1
+ * req: |-----| (bo_offset=n)
+ *
+ * 0 b 1 a' 2
+ * new: |-----|-----| (b.bo_offset=n, a.bo_offset=n+1)
+ *
+ * .. note::
+ * We expect to see the same result for a request with a different BO
+ * and/or non-contiguous BO offset.
+ *
+ *
+ * 6) Existent mapping is a superset of the requested mapping. Split it up, but
+ * indicate that the backing PTEs could be kept.
+ *
+ * ::
+ *
+ * 0 a 2
+ * old: |-----------| (bo_offset=n)
+ *
+ * 0 a 1
+ * req: |-----| (bo_offset=n)
+ *
+ * 0 a 1 a' 2
+ * new: |-----|-----| (a.bo_offset=n, a'.bo_offset=n+1)
+ *
+ *
+ * 7) Requested mapping's range is a right aligned subset of the existent one,
+ * but backed by a different BO. Hence, map the requested mapping and split
+ * the existent one, without adjusting the BO offset.
+ *
+ * ::
+ *
+ * 0 a 2
+ * old: |-----------| (bo_offset=n)
+ *
+ * 1 b 2
+ * req: |-----| (bo_offset=m)
+ *
+ * 0 a 1 b 2
+ * new: |-----|-----| (a.bo_offset=n,b.bo_offset=m)
+ *
+ *
+ * 8) Existent mapping is a superset of the requested mapping. Split it up, but
+ * indicate that the backing PTEs could be kept.
+ *
+ * ::
+ *
+ * 0 a 2
+ * old: |-----------| (bo_offset=n)
+ *
+ * 1 a 2
+ * req: |-----| (bo_offset=n+1)
+ *
+ * 0 a' 1 a 2
+ * new: |-----|-----| (a'.bo_offset=n, a.bo_offset=n+1)
+ *
+ *
+ * 9) Existent mapping is overlapped at the end by the requested mapping backed
+ * by a different BO. Hence, map the requested mapping and split up the
+ * existent one, without adjusting the BO offset.
+ *
+ * ::
+ *
+ * 0 a 2
+ * old: |-----------| (bo_offset=n)
+ *
+ * 1 b 3
+ * req: |-----------| (bo_offset=m)
+ *
+ * 0 a 1 b 3
+ * new: |-----|-----------| (a.bo_offset=n,b.bo_offset=m)
+ *
+ *
+ * 10) Existent mapping is overlapped by the requested mapping, both having the
+ * same backing BO with a contiguous offset. Indicate the backing PTEs of
+ * the old mapping could be kept.
+ *
+ * ::
+ *
+ * 0 a 2
+ * old: |-----------| (bo_offset=n)
+ *
+ * 1 a 3
+ * req: |-----------| (bo_offset=n+1)
+ *
+ * 0 a' 1 a 3
+ * new: |-----|-----------| (a'.bo_offset=n, a.bo_offset=n+1)
+ *
+ *
+ * 11) Requested mapping's range is a centered subset of the existent one
+ * having a different backing BO. Hence, map the requested mapping and split
+ * up the existent one in two mappings, adjusting the BO offset of the right
+ * one accordingly.
+ *
+ * ::
+ *
+ * 0 a 3
+ * old: |-----------------| (bo_offset=n)
+ *
+ * 1 b 2
+ * req: |-----| (bo_offset=m)
+ *
+ * 0 a 1 b 2 a' 3
+ * new: |-----|-----|-----| (a.bo_offset=n,b.bo_offset=m,a'.bo_offset=n+2)
+ *
+ *
+ * 12) Requested mapping is a contiguous subset of the existent one. Split it
+ * up, but indicate that the backing PTEs could be kept.
+ *
+ * ::
+ *
+ * 0 a 3
+ * old: |-----------------| (bo_offset=n)
+ *
+ * 1 a 2
+ * req: |-----| (bo_offset=n+1)
+ *
+ * 0 a' 1 a 2 a'' 3
+ * old: |-----|-----|-----| (a'.bo_offset=n, a.bo_offset=n+1, a''.bo_offset=n+2)
+ *
+ *
+ * 13) Existent mapping is a right aligned subset of the requested one, hence
+ * replace the existent one.
+ *
+ * ::
+ *
+ * 1 a 2
+ * old: |-----| (bo_offset=n+1)
+ *
+ * 0 a 2
+ * req: |-----------| (bo_offset=n)
+ *
+ * 0 a 2
+ * new: |-----------| (bo_offset=n)
+ *
+ * .. note::
+ * We expect to see the same result for a request with a different bo
+ * and/or non-contiguousbo_offset.
+ *
+ *
+ * 14) Existent mapping is a centered subset of the requested one, hence
+ * replace the existent one.
+ *
+ * ::
+ *
+ * 1 a 2
+ * old: |-----| (bo_offset=n+1)
+ *
+ * 0 a 3
+ * req: |----------------| (bo_offset=n)
+ *
+ * 0 a 3
+ * new: |----------------| (bo_offset=n)
+ *
+ * .. note::
+ * We expect to see the same result for a request with a different bo
+ * and/or non-contiguousbo_offset.
+ *
+ *
+ * 15) Existent mappings is overlapped at the beginning by the requested mapping
+ * backed by a different BO. Hence, map the requested mapping and split up
+ * the existent one, adjusting it's BO offset accordingly.
+ *
+ * ::
+ *
+ * 1 a 3
+ * old: |-----------| (bo_offset=n)
+ *
+ * 0 b 2
+ * req: |-----------| (bo_offset=m)
+ *
+ * 0 b 2 a' 3
+ * new: |-----------|-----| (b.bo_offset=m,a.bo_offset=n+2)
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +5
-0+ +/** +* DOC: Locking +* +* Generally, the GPU VA manager does not take care of locking itself, it is
+ * the drivers responsibility to take care about locking. Drivers might want to
+ * protect the following operations: inserting, removing and iterating
+ *&drm_gpuvaobjects as well as generating all kinds of operations, such as
+ * split / merge or prefetch.
+ *
+ * The GPU VA manager also does not take care of the locking of the backing
+ *&drm_gem_objectbuffers GPU VA lists by itself; drivers areQuoted 8 lines
Quoted 6 lines
responsible to
+ * enforce mutual exclusion using either the GEMsdma_resvlock or alternatively
+ * a driver specific external lock by setting the @DRM_GPUVA_MANAGER_LOCK_EXTERN
+ * flag.Is the external lock used or anticipated by any WIP implementation? If not I suggest to leave it out until there is a user.
Quoted 5 lines
I think the PowerVR driver requires this. However, it's generally useful for any driver using direct callbacks rather than
drm_gpuva_ops.Once the page table handling in Nouveau is re-worked, and direct callbacks can be used, I probably want to use this in Nouveau as well.
Gonna fix up all other comments.
Sounds good to me. The main concern here with the external lock was testing coverage if no user. I don't have anything against it per se.
Going on vacation on Monday, but that Acked-by: once addressed on IRC still holds.
Thanks,
Thomas
Quoted 2,410 lines
- Danilo
Quoted 2,407 lines
Quoted 224 lines
+ *
+ * For the latter, functions such asdrm_gpuva_link()ordrm_gpuva_unlink()
+ * contain lockdep checks to indicate locking issues. For this to work drivers
+ * must provide (in case the @DRM_GPUVA_MANAGER_LOCK_EXTERN flag is set) their
+ * external lock withdrm_gpuva_manager_set_ext_lock()after initialization.
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +5
-0+ +/** +* DOC: Examples +* +* This section gives two examples on how to let the DRM GPUVA Manager generate
+ *&drm_gpuva_opin order to satisfy a given map or unmap request and how to
+ * make use of them.
+ *
+ * The below code is strictly limited to illustrate the generic usage pattern.
+ * To maintain simplicitly, it doesn't make use of any abstractions for common
+ * code, different (asyncronous) stages with fence signalling critical paths,
+ * any other helpers or error handling in terms of freeing memory and dropping
+ * previously taken locks.
+ *
+ * 1) Obtain a list of&drm_gpuva_opto create a new mapping::
+ *
+ * // Allocates a new&drm_gpuva.
+ * structdrm_gpuva*driver_gpuva_alloc(void);
+ *
+ * // Typically drivers would embedd the&drm_gpuva_managerand&drm_gpuva
+ * // structure in individual driver structures and lock the dma-resv with
+ * //drm_execor similar helpers.
+ * int driver_mapping_create(structdrm_gpuva_manager*mgr,
+ * u64 addr, u64 range,
+ * structdrm_gem_object*obj, u64 offset)
+ * {
+ * structdrm_gpuva_ops*ops;
+ * structdrm_gpuva_op*op
+ *
+ *driver_lock_va_space();
+ * ops = drm_gpuva_sm_map_ops_create(mgr, addr, range,
+ * obj, offset);
+ * if (IS_ERR(ops))
+ * returnPTR_ERR(ops);
+ *
+ *drm_gpuva_for_each_op(op, ops){
+ * structdrm_gpuva*va;
+ *
+ * switch (op->op) {
+ * caseDRM_GPUVA_OP_MAP:
+ * va =driver_gpuva_alloc();
+ * if (!va)
+ * ; // unwind previous VA space updates,
+ * // free memory and unlock
+ *
+ *driver_vm_map();
+ *drm_gpuva_map(mgr, va, &op->map);
+ *drm_gpuva_link(va);
+ *
+ * break;
+ * caseDRM_GPUVA_OP_REMAP: {
+ * structdrm_gpuva*prev = NULL, *next = NULL;
+ *
+ * va =op->remap.unmap->va;
+ *
+ * if (op->remap.prev) {
+ * prev =driver_gpuva_alloc();
+ * if (!prev)
+ * ; // unwind previous VA space
+ * // updates, free memory and
+ * // unlock
+ * }
+ *
+ * if (op->remap.next) {
+ * next =driver_gpuva_alloc();
+ * if (!next)
+ * ; // unwind previous VA space
+ * // updates, free memory and
+ * // unlock
+ * }
+ *
+ *driver_vm_remap();
+ *drm_gpuva_remap(prev, next, &op->remap);
+ *
+ *drm_gpuva_unlink(va);
+ * if (prev)
+ *drm_gpuva_link(prev);
+ * if (next)
+ *drm_gpuva_link(next);
+ *
+ * break;
+ * }
+ * caseDRM_GPUVA_OP_UNMAP:
+ * va =op->unmap->va;
+ *
+ *driver_vm_unmap();
+ *drm_gpuva_unlink(va);
+ *drm_gpuva_unmap(&op->unmap);
+ *
+ * break;
+ * default:
+ * break;
+ * }
+ * }
+ *driver_unlock_va_space();
+ *
+ * return 0;
+ * }
+ *
+ * 2) Receive a callback for each&drm_gpuva_opto create a new mapping::
+ *
+ * structdriver_context{
+ * structdrm_gpuva_manager*mgr;
+ * structdrm_gpuva*new_va;
+ * structdrm_gpuva*prev_va;
+ * structdrm_gpuva*next_va;
+ * };
+ *
+ * // ops to pass todrm_gpuva_manager_init()
+ * static const structdrm_gpuva_fn_opsdriver_gpuva_ops= {
+ * .sm_step_map =driver_gpuva_map,
+ * .sm_step_remap =driver_gpuva_remap,
+ * .sm_step_unmap =driver_gpuva_unmap,
+ * };
+ *
+ * // Typically drivers would embedd the&drm_gpuva_managerand&drm_gpuva
+ * // structure in individual driver structures and lock the dma-resv with
+ * //drm_execor similar helpers.
+ * int driver_mapping_create(structdrm_gpuva_manager*mgr,
+ * u64 addr, u64 range,
+ * structdrm_gem_object*obj, u64 offset)
+ * {
+ * structdriver_contextctx;
+ * structdrm_gpuva_ops*ops;
+ * structdrm_gpuva_op*op;
+ * int ret = 0;
+ *
+ * ctx.mgr = mgr;
+ *
+ * ctx.new_va =kzalloc(sizeof(*ctx.new_va), GFP_KERNEL);
+ * ctx.prev_va =kzalloc(sizeof(*ctx.prev_va), GFP_KERNEL);
+ * ctx.next_va =kzalloc(sizeof(*ctx.next_va), GFP_KERNEL);
+ * if (!ctx.new_va || !ctx.prev_va || !ctx.next_va) {
+ * ret =-ENOMEM;
+ * goto out;
+ * }
+ *
+ *driver_lock_va_space();
+ * ret =drm_gpuva_sm_map(mgr, &ctx, addr, range, obj, offset);
+ *driver_unlock_va_space();
+ *
+ * out:
+ *kfree(ctx.new_va);
+ *kfree(ctx.prev_va);
+ *kfree(ctx.next_va);
+ * return ret;
+ * }
+ *
+ * intdriver_gpuva_map(struct drm_gpuva_op *op, void *__ctx)
+ * {
+ * structdriver_context*ctx =__ctx;
+ *
+ *drm_gpuva_map(ctx->mgr, ctx->new_va, &op->map);
+ *
+ *drm_gpuva_link(ctx->new_va);
+ *
+ * // prevent the new GPUVA from being freed in
+ * //driver_mapping_create()
+ *ctx->new_va= NULL;
+ *
+ * return 0;
+ * }
+ *
+ * intdriver_gpuva_remap(struct drm_gpuva_op *op, void *__ctx)
+ * {
+ * structdriver_context*ctx =__ctx;
+ *
+ *drm_gpuva_remap(ctx->prev_va, ctx->next_va, &op->remap);
+ *
+ *drm_gpuva_unlink(op->remap.unmap->va);
+ *kfree(op->remap.unmap->va);
+ *
+ * if (op->remap.prev) {
+ *drm_gpuva_link(ctx->prev_va);
+ *ctx->prev_va= NULL;
+ * }
+ *
+ * if (op->remap.next) {
+ *drm_gpuva_link(ctx->next_va);
+ *ctx->next_va= NULL;
+ * }
+ *
+ * return 0;
+ * }
+ *
+ * intdriver_gpuva_unmap(struct drm_gpuva_op *op, void *__ctx)
+ * {
+ *drm_gpuva_unlink(op->unmap.va);
+ *drm_gpuva_unmap(&op->unmap);
+ *kfree(op->unmap.va);
+ *
+ * return 0;
+ * }
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +2
-0+ +#define to_drm_gpuva (__node) container_of((__node), struct
drm_gpuva, rb.node)drivers/gpu/drm/drm_gpuva_mgr.c new file +5
-0+ +#define GPUVA_START (node) ((node)->va.addr) +#define GPUVA_LAST (node) ((node)->va.addr + (node)->va.range - 1) + +/* We do not actually use drm_gpuva_it_next(), tell the compiler to Quoted 6 lines
not complain
+ * about this.
+ */
+INTERVAL_TREE_DEFINE(structdrm_gpuva, rb.node, u64, rb.__subtree_last,
+GPUVA_START,GPUVA_LAST, static __attribute__((unused)),Would s/__attribute__((unused))/__maybe_unused/ work here?
drivers/gpu/drm/drm_gpuva_mgr.c line 617
Quoted 7 lines
+
drm_gpuva_it)drivers/gpu/drm/drm_gpuva_mgr.c new file +6
-0+ +static int __drm_gpuva_insert (struct drm_gpuva_manager *mgr, + struct drm_gpuva *va); +static void __drm_gpuva_remove (struct drm_gpuva *va); + +static inline bool"static inline" is typically used only in header files, since the compiler should be smart enough to inline if beneficial. Prefer "static". Also in multiple places below.
Quoted 33 lines
drivers/gpu/drm/drm_gpuva_mgr.c new file +10
-0+drm_gpuva_check_overflow(u64 addr, u64 range) +{ + u64 end; + + return WARN(check_add_overflow(addr, range, &end), + "GPUVA address limited to %lu bytes.\n", sizeof(end)); +} + +static inline bool +drm_gpuva_in_mm_range(struct drm_gpuva_manager *mgr, u64 addr, u64range)
drivers/gpu/drm/drm_gpuva_mgr.c new file +10
-0+{ + u64 end = addr + range; + u64 mm_start = mgr->mm_start; + u64 mm_end = mm_start + mgr->mm_range; + + return addr >= mm_start && end <= mm_end; +} + +static inline bool +drm_gpuva_in_kernel_node(struct drm_gpuva_manager *mgr, u64 addr,u64 range)
drivers/gpu/drm/drm_gpuva_mgr.c new file +11
-0+{ + u64 end = addr + range; + u64 kstart = mgr->kernel_alloc_node.va.addr; + u64 krange = mgr->kernel_alloc_node.va.range; + u64 kend = kstart + krange; + + return krange && addr < kend && kstart < end; +} + +static inline bool +drm_gpuva_range_valid(struct drm_gpuva_manager *mgr,drivers/gpu/drm/drm_gpuva_mgr.c line 658
Quoted 10 lines
drivers/gpu/drm/drm_gpuva_mgr.c new file +10
-0+ u64 addr, u64 range) +{ + + return !drm_gpuva_check_overflow(addr, range) && + drm_gpuva_in_mm_range(mgr, addr, range) && + !drm_gpuva_in_kernel_node(mgr, addr, range); +} + +/** +* drm_gpuva_manager_init - initialize a &drm_gpuva_manager Function kerneldoc names should end with "()":
drm_gpuva_manager_init(). General comment across the patch.Quoted 1,220 lines
+ * @mgr: pointer to the
&drm_gpuva_managerto initialize
+ * @name: the name of the GPU VA space
+ * @start_offset: the start offset of the GPU VA space
+ * @range: the size of the GPU VA space
+ * @reserve_offset: the start of the kernel reserved GPU VA area
+ * @reserve_range: the size of the kernel reserved GPU VA area
+ * @ops:&drm_gpuva_fn_opscalled on&drm_gpuva_sm_map/&drm_gpuva_sm_unmap
+ * @flags: the feature flags for the&drm_gpuva_manager
+ *
+ * The&drm_gpuva_managermust be initialized with this function before use.
+ *
+ * Note that @mgr must be cleared to 0 before calling this function. The given
+ * &name is expected to be managed by the surrounding driver structures.
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +89
-0+void +drm_gpuva_manager_init (struct drm_gpuva_manager *mgr, + const char *name, + u64 start_offset, u64 range, + u64 reserve_offset, u64 reserve_range, + const struct drm_gpuva_fn_ops *ops, + enum drm_gpuva_manager_flags flags) +{ + mgr->rb.tree = RB_ROOT_CACHED; + INIT_LIST_HEAD(&mgr->rb.list); + + drm_gpuva_check_overflow(start_offset, range); + mgr->mm_start = start_offset; + mgr->mm_range = range; + + mgr->name = name ? name : "unknown"; + mgr->flags = flags; + mgr->ops = ops; + + memset(&mgr->kernel_alloc_node, 0, sizeof(struct drm_gpuva)); + + if (reserve_range) { + mgr->kernel_alloc_node.va.addr = reserve_offset; + mgr->kernel_alloc_node.va.range = reserve_range; + + if (likely(!drm_gpuva_check_overflow(reserve_offset, + reserve_range))) + __drm_gpuva_insert(mgr, &mgr->kernel_alloc_node); + } + +} +EXPORT_SYMBOL (drm_gpuva_manager_init); + +/** +* drm_gpuva_manager_destroy - cleanup a &drm_gpuva_manager +* @mgr: pointer to the &drm_gpuva_manager to clean up +* +* Note that it is a bug to call this function on a manager that still +* holds GPU VA mappings. +*/ +void +drm_gpuva_manager_destroy (struct drm_gpuva_manager *mgr) +{ + mgr->name = NULL; + + if (mgr->kernel_alloc_node.va.range) + __drm_gpuva_remove(&mgr->kernel_alloc_node); + + WARN(!RB_EMPTY_ROOT(&mgr->rb.tree.rb_root), + "GPUVA tree is not empty, potentially leaking memory."); +} +EXPORT_SYMBOL (drm_gpuva_manager_destroy); + +static int +__drm_gpuva_insert (struct drm_gpuva_manager *mgr, + struct drm_gpuva *va) +{ + struct rb_node *node; + struct list_head *head; + + if (drm_gpuva_it_iter_first(&mgr->rb.tree, + GPUVA_START(va), + GPUVA_LAST(va))) + return -EEXIST; + + va->mgr = mgr; + + drm_gpuva_it_insert(va, &mgr->rb.tree); + + node = rb_prev(&va->rb.node); + if (node) + head = &(to_drm_gpuva(node))->rb.entry; + else + head = &mgr->rb.list; + + list_add(&va->rb.entry, head); + + return 0; +} + +/** +* drm_gpuva_insert - insert a &drm_gpuva +* @mgr: the &drm_gpuva_manager to insert the &drm_gpuva in +* @va: the &drm_gpuva to insert +* +* Insert a &drm_gpuva with a given address and range into a +* &drm_gpuva_manager. +* +* It is safe to use this function using the safe versions of iterating the GPU
+ * VA space, such asdrm_gpuva_for_each_va_safe()and
+ *drm_gpuva_for_each_va_range_safe().
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +28
-0+int +drm_gpuva_insert (struct drm_gpuva_manager *mgr, + struct drm_gpuva *va) +{ + u64 addr = va->va.addr; + u64 range = va->va.range; + + if (unlikely(!drm_gpuva_range_valid(mgr, addr, range))) + return -EINVAL; + + return __drm_gpuva_insert(mgr, va); +} +EXPORT_SYMBOL (drm_gpuva_insert); + +static void +__drm_gpuva_remove (struct drm_gpuva *va) +{ + drm_gpuva_it_remove(va, &va->mgr->rb.tree); + list_del_init(&va->rb.entry); +} + +/** +* drm_gpuva_remove - remove a &drm_gpuva +* @va: the &drm_gpuva to remove +* +* This removes the given &va from the underlaying tree. +* +* It is safe to use this function using the safe versions of iterating the GPU
+ * VA space, such asdrm_gpuva_for_each_va_safe()and
+ *drm_gpuva_for_each_va_range_safe().
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +19
-0+void +drm_gpuva_remove (struct drm_gpuva *va) +{ + struct drm_gpuva_manager *mgr = va->mgr; + + if (unlikely(va == &mgr->kernel_alloc_node)) { + WARN(1, "Can't destroy kernel reserved node.\n"); + return; + } + + __drm_gpuva_remove(va); +} +EXPORT_SYMBOL (drm_gpuva_remove); + +/** +* drm_gpuva_link - link a &drm_gpuva +* @va: the &drm_gpuva to link +* +* This adds the given &va to the GPU VA list of the
&drm_gem_objectit is
+ * associated with.
+ *
+ * This function expects the caller to protect the GEM's GPUVA list against
+ * concurrent access using the GEMsdma_resvlock.
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +23
-0+void +drm_gpuva_link (struct drm_gpuva *va) +{ + struct drm_gpuva_manager *mgr = va->mgr; + struct drm_gem_object *obj = va->gem.obj; + + if (unlikely(!obj)) + return; + + if (drm_gpuva_manager_external_lock(mgr)) + drm_gpuva_manager_ext_assert_held(mgr); + else + dma_resv_assert_held(obj->resv); + + list_add_tail(&va->gem.entry, &obj->gpuva.list); +} +EXPORT_SYMBOL (drm_gpuva_link); + +/** +* drm_gpuva_unlink - unlink a &drm_gpuva +* @va: the &drm_gpuva to unlink +* +* This removes the given &va from the GPU VA list of the
&drm_gem_objectit is
+ * associated with.
+ *
+ * This function expects the caller to protect the GEM's GPUVA list against
+ * concurrent access using the GEMsdma_resvlock.
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +71
-0+void +drm_gpuva_unlink (struct drm_gpuva *va) +{ + struct drm_gpuva_manager *mgr = va->mgr; + struct drm_gem_object *obj = va->gem.obj; + + if (unlikely(!obj)) + return; + + if (drm_gpuva_manager_external_lock(mgr)) + drm_gpuva_manager_ext_assert_held(mgr); + else + dma_resv_assert_held(obj->resv); + + list_del_init(&va->gem.entry); +} +EXPORT_SYMBOL (drm_gpuva_unlink); + +/** +* drm_gpuva_find_first - find the first &drm_gpuva in the given range +* @mgr: the &drm_gpuva_manager to search in +* @addr: the &drm_gpuvas address +* @range: the &drm_gpuvas range +* +* Returns: the first &drm_gpuva within the given range +*/ +struct drm_gpuva * +drm_gpuva_find_first (struct drm_gpuva_manager *mgr, + u64 addr, u64 range) +{ + u64 last = addr + range - 1; + + return drm_gpuva_it_iter_first(&mgr->rb.tree, addr, last); +} +EXPORT_SYMBOL (drm_gpuva_find_first); + +/** +* drm_gpuva_find - find a &drm_gpuva +* @mgr: the &drm_gpuva_manager to search in +* @addr: the &drm_gpuvas address +* @range: the &drm_gpuvas range +* +* Returns: the &drm_gpuva at a given &addr and with a given &range +*/ +struct drm_gpuva * +drm_gpuva_find (struct drm_gpuva_manager *mgr, + u64 addr, u64 range) +{ + struct drm_gpuva *va; + + va = drm_gpuva_find_first(mgr, addr, range); + if (!va) + goto out; + + if (va->va.addr != addr || + va->va.range != range) + goto out; + + return va; + +out: + return NULL; +} +EXPORT_SYMBOL (drm_gpuva_find); + +/** +* drm_gpuva_find_prev - find the &drm_gpuva before the given address +* @mgr: the &drm_gpuva_manager to search in +* @start: the given GPU VA's start address +* +* Find the adjacent &drm_gpuva before the GPU VA with given &start address.
+ *
+ * Note that if there is any free space between the GPU VA mappings no mapping
+ * is returned.
+ *
+ * Returns: a pointer to the found&drm_gpuvaor NULL if none was found
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +16
-0+struct drm_gpuva * +drm_gpuva_find_prev (struct drm_gpuva_manager *mgr,u64 start) +{ + if (!drm_gpuva_range_valid(mgr, start - 1, 1)) + return NULL; + + return drm_gpuva_it_iter_first(&mgr->rb.tree, start - 1, start); +} +EXPORT_SYMBOL (drm_gpuva_find_prev); + +/** +* drm_gpuva_find_next - find the &drm_gpuva after the given address +* @mgr: the &drm_gpuva_manager to search in +* @end: the given GPU VA's end address +* +* Find the adjacent &drm_gpuva after the GPU VA with given &end address.
+ *
+ * Note that if there is any free space between the GPU VA mappings no mapping
+ * is returned.
+ *
+ * Returns: a pointer to the found&drm_gpuvaor NULL if none was found
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +12
-0+struct drm_gpuva * +drm_gpuva_find_next (struct drm_gpuva_manager *mgr,u64 end) +{ + if (!drm_gpuva_range_valid(mgr, end, 1)) + return NULL; + + return drm_gpuva_it_iter_first(&mgr->rb.tree, end, end + 1); +} +EXPORT_SYMBOL (drm_gpuva_find_next); + +/** +* drm_gpuva_interval_empty - indicate whether a given interval of the VA space
+ * is empty
+ * @mgr: the&drm_gpuva_managerto check the range for
+ * @addr: the start address of the range
+ * @range: the range of the interval
+ *
+ * Returns: true if the interval is empty, false otherwise
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +2
-0+bool +drm_gpuva_interval_empty(struct drm_gpuva_manager *mgr, u64 addr,u64 range)
drivers/gpu/drm/drm_gpuva_mgr.c new file +13
-0+{ + return !drm_gpuva_find_first(mgr, addr, range); +} +EXPORT_SYMBOL (drm_gpuva_interval_empty); + +/** +* drm_gpuva_map - helper to insert a &drm_gpuva according to a +* &drm_gpuva_op_map +* @mgr: the &drm_gpuva_manager +* @va: the &drm_gpuva to insert +* @op: the &drm_gpuva_op_map to initialize @va with +* +* Initializes the @va from the @op and inserts it into the given @mgr.
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +18
-0+void +drm_gpuva_map (struct drm_gpuva_manager *mgr, + struct drm_gpuva *va, + struct drm_gpuva_op_map *op) +{ + drm_gpuva_init_from_op(va, op); + drm_gpuva_insert(mgr, va); +} +EXPORT_SYMBOL (drm_gpuva_map); + +/** +* drm_gpuva_remap - helper to remap a &drm_gpuva according to a +* &drm_gpuva_op_remap +* @prev: the &drm_gpuva to remap when keeping the start of a mapping +* @next: the &drm_gpuva to remap when keeping the end of a mapping +* @op: the &drm_gpuva_op_remap to initialize @prev and @next with +* +* Removes the currently mapped &drm_gpuva and remaps it using @prev and/or
+ * @next.
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +98
-0+void +drm_gpuva_remap (struct drm_gpuva *prev, + struct drm_gpuva *next, + struct drm_gpuva_op_remap *op) +{ + struct drm_gpuva *curr = op->unmap->va; + struct drm_gpuva_manager *mgr = curr->mgr; + struct drm_gpuva_op_map *map; + + drm_gpuva_remove(curr); + + if ((map = op->prev)) { + drm_gpuva_init_from_op(prev, map); + drm_gpuva_insert(mgr, prev); + } + + if ((map = op->next)) { + drm_gpuva_init_from_op(next, map); + drm_gpuva_insert(mgr, next); + } +} +EXPORT_SYMBOL (drm_gpuva_remap); + +/** +* drm_gpuva_unmap - helper to remove a &drm_gpuva according to a +* &drm_gpuva_op_unmap +* @op: the &drm_gpuva_op_unmap specifying the &drm_gpuva to remove +* +* Removes the &drm_gpuva associated with the &drm_gpuva_op_unmap. +*/ +void +drm_gpuva_unmap (struct drm_gpuva_op_unmap *op) +{ + drm_gpuva_remove(op->va); +} +EXPORT_SYMBOL (drm_gpuva_unmap); + +static int +op_map_cb (const struct drm_gpuva_fn_ops *fn,void *priv, + u64 addr, u64 range, + struct drm_gem_object *obj, u64 offset) +{ + struct drm_gpuva_op op = {}; + + op.op = DRM_GPUVA_OP_MAP; + op.map.va.addr = addr; + op.map.va.range = range; + op.map.gem.obj = obj; + op.map.gem.offset = offset; + + return fn->sm_step_map(&op, priv); +} + +static int +op_remap_cb (const struct drm_gpuva_fn_ops *fn,void *priv, + struct drm_gpuva_op_map *prev, + struct drm_gpuva_op_map *next, + struct drm_gpuva_op_unmap *unmap) +{ + struct drm_gpuva_op op = {}; + struct drm_gpuva_op_remap *r; + + op.op = DRM_GPUVA_OP_REMAP; + r = &op.remap; + r->prev = prev; + r->next = next; + r->unmap = unmap; + + return fn->sm_step_remap(&op, priv); +} + +static int +op_unmap_cb (const struct drm_gpuva_fn_ops *fn,void *priv, + struct drm_gpuva *va, bool merge) +{ + struct drm_gpuva_op op = {}; + + op.op = DRM_GPUVA_OP_UNMAP; + op.unmap.va = va; + op.unmap.keep = merge; + + return fn->sm_step_unmap(&op, priv); +} + +static int +__drm_gpuva_sm_map (struct drm_gpuva_manager *mgr, + const struct drm_gpuva_fn_ops *ops, void *priv, + u64 req_addr, u64 req_range, + struct drm_gem_object *req_obj, u64 req_offset) +{ + struct drm_gpuva *va, *next, *prev = NULL; + u64 req_end = req_addr + req_range; + int ret; + + if (unlikely(!drm_gpuva_range_valid(mgr, req_addr, req_range))) + return -EINVAL; + + drm_gpuva_for_each_va_range_safe(va, next, mgr, req_addr,
req_end) {
+ structdrm_gem_object*obj =va->gem.obj; + u64 offset =va->gem.offset;
+ u64 addr =va->va.addr;
+ u64 range =va->va.range;
+ u64 end = addr + range;
+ bool merge = !!va->gem.obj;drivers/gpu/drm/drm_gpuva_mgr.c new file +137
-0+ + if (addr == req_addr) { + merge &= obj == req_obj && + offset == req_offset; + + if (end == req_end) { + ret = op_unmap_cb(ops, priv, va, merge); + if (ret) + return ret; + break; + } + + if (end < req_end) { + ret = op_unmap_cb(ops, priv, va, merge); + if (ret) + return ret; + goto next; + } + + if (end > req_end) { + struct drm_gpuva_op_map n = { + .va.addr = req_end, + .va.range = range - req_range, + .gem.obj = obj, + .gem.offset = offset + req_range, + }; + struct drm_gpuva_op_unmap u = { + .va = va, + .keep = merge, + }; + + ret = op_remap_cb(ops, priv, NULL, &n, &u); + if (ret) + return ret; + break; + } + } else if (addr < req_addr) { + u64 ls_range = req_addr - addr; + struct drm_gpuva_op_map p = { + .va.addr = addr, + .va.range = ls_range, + .gem.obj = obj, + .gem.offset = offset, + }; + struct drm_gpuva_op_unmap u = { .va = va }; + + merge &= obj == req_obj && + offset + ls_range == req_offset; + u.keep = merge; + + if (end == req_end) { + ret = op_remap_cb(ops, priv, &p, NULL, &u); + if (ret) + return ret; + break; + } + + if (end < req_end) { + ret = op_remap_cb(ops, priv, &p, NULL, &u); + if (ret) + return ret; + goto next; + } + + if (end > req_end) { + struct drm_gpuva_op_map n = { + .va.addr = req_end, + .va.range = end - req_end, + .gem.obj = obj, + .gem.offset = offset + ls_range + + req_range, + }; + + ret = op_remap_cb(ops, priv, &p, &n, &u); + if (ret) + return ret; + break; + } + } else if (addr > req_addr) { + merge &= obj == req_obj && + offset == req_offset + + (addr - req_addr); + + if (end == req_end) { + ret = op_unmap_cb(ops, priv, va, merge); + if (ret) + return ret; + break; + } + + if (end < req_end) { + ret = op_unmap_cb(ops, priv, va, merge); + if (ret) + return ret; + goto next; + } + + if (end > req_end) { + struct drm_gpuva_op_map n = { + .va.addr = req_end, + .va.range = end - req_end, + .gem.obj = obj, + .gem.offset = offset + req_end - addr, + }; + struct drm_gpuva_op_unmap u = { + .va = va, + .keep = merge, + }; + + ret = op_remap_cb(ops, priv, NULL, &n, &u); + if (ret) + return ret; + break; + } + } +next: + prev = va; + } + + return op_map_cb(ops, priv, + req_addr, req_range, + req_obj, req_offset); +} + +static int +__drm_gpuva_sm_unmap (struct drm_gpuva_manager *mgr, + const struct drm_gpuva_fn_ops *ops, void *priv, + u64 req_addr, u64 req_range) +{ + struct drm_gpuva *va, *next; + u64 req_end = req_addr + req_range; + int ret; + + if (unlikely(!drm_gpuva_range_valid(mgr, req_addr, req_range))) + return -EINVAL; + + drm_gpuva_for_each_va_range_safe(va, next, mgr, req_addr,
req_end) {
+ structdrm_gpuva_op_mapprev = {}, next = {}; + boolprev_split= false,next_split= false; + structdrm_gem_object*obj =va->gem.obj; + u64 offset =va->gem.offset;
+ u64 addr =va->va.addr;
+ u64 range =va->va.range;
+ u64 end = addr + range;drivers/gpu/drm/drm_gpuva_mgr.c new file +48
-0+ + if (addr < req_addr) { + prev.va.addr = addr; + prev.va.range = req_addr - addr; + prev.gem.obj = obj; + prev.gem.offset = offset; + + prev_split = true; + } + + if (end > req_end) { + next.va.addr = req_end; + next.va.range = end - req_end; + next.gem.obj = obj; + next.gem.offset = offset + (req_end - addr); + + next_split = true; + } + + if (prev_split || next_split) { + struct drm_gpuva_op_unmap unmap = { .va = va }; + + ret = op_remap_cb(ops, priv, + prev_split ? &prev : NULL, + next_split ? &next : NULL, + &unmap); + if (ret) + return ret; + } else { + ret = op_unmap_cb(ops, priv, va, false); + if (ret) + return ret; + } + } + + return 0; +} + +/** +* drm_gpuva_sm_map - creates the &drm_gpuva_op split/merge steps +* @mgr: the &drm_gpuva_manager representing the GPU VA space +* @req_addr: the start address of the new mapping +* @req_range: the range of the new mapping +* @req_obj: the &drm_gem_object to map +* @req_offset: the offset within the &drm_gem_object +* @priv: pointer to a driver private data structure +* +* This function iterates the given range of the GPU VA space. It utilizes the
+ *&drm_gpuva_fn_opsto call back into the driver providing the split and merge
+ * steps.
+ *
+ * Drivers may use these callbacks to update the GPU VA space right away within
+ * the callback. In case the driver decides to copy and store the operations for
+ * later processing neither this function nor&drm_gpuva_sm_unmapis allowed to
+ * be called before the &drm_gpuva_manager's view of the GPU VA space was
+ * updated with the previous set of operations. To update the
+ * &drm_gpuva_manager's view of the GPU VA spacedrm_gpuva_insert(),
+ *drm_gpuva_destroy_locked()and/ordrm_gpuva_destroy_unlocked()should be
+ * used.
+ *
+ * A sequence of callbacks can contain map, unmap and remap operations, but
+ * the sequence of callbacks might also be empty if no operation is required,
+ * e.g. if the requested mapping already exists in the exact same way.
+ *
+ * There can be an arbitrary amount of unmap operations, a maximum of two remap
+ * operations and a single map operation. The latter one represents the original
+ * map operation requested by the caller.
+ *
+ * Returns: 0 on success or a negative error code
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +26
-0+int +drm_gpuva_sm_map (struct drm_gpuva_manager *mgr,void *priv, + u64 req_addr, u64 req_range, + struct drm_gem_object *req_obj, u64 req_offset) +{ + const struct drm_gpuva_fn_ops *ops = mgr->ops; + + if (unlikely(!(ops && ops->sm_step_map && + ops->sm_step_remap && + ops->sm_step_unmap))) + return -EINVAL; + + return __drm_gpuva_sm_map(mgr, ops, priv, + req_addr, req_range, + req_obj, req_offset); +} +EXPORT_SYMBOL (drm_gpuva_sm_map); + +/** +* drm_gpuva_sm_unmap - creates the &drm_gpuva_ops to split on unmap +* @mgr: the &drm_gpuva_manager representing the GPU VA space +* @priv: pointer to a driver private data structure +* @req_addr: the start address of the range to unmap +* @req_range: the range of the mappings to unmap +* +* This function iterates the given range of the GPU VA space. It utilizes the
+ *&drm_gpuva_fn_opsto call back into the driver providing the operations to
+ * unmap and, if required, split existent mappings.
+ *
+ * Drivers may use these callbacks to update the GPU VA space right away within
+ * the callback. In case the driver decides to copy and store the operations for
+ * later processing neither this function nor&drm_gpuva_sm_mapis allowed to be
+ * called before the &drm_gpuva_manager's view of the GPU VA space was updated
+ * with the previous set of operations. To update the &drm_gpuva_manager's view
+ * of the GPU VA spacedrm_gpuva_insert(),drm_gpuva_destroy_locked()and/or
+ *drm_gpuva_destroy_unlocked()should be used.
+ *
+ * A sequence of callbacks can contain unmap and remap operations, depending on
+ * whether there are actual overlapping mappings to split.
+ *
+ * There can be an arbitrary amount of unmap operations and a maximum of two
+ * remap operations.
+ *
+ * Returns: 0 on success or a negative error code
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +108
-0+int +drm_gpuva_sm_unmap (struct drm_gpuva_manager *mgr,void *priv, + u64 req_addr, u64 req_range) +{ + const struct drm_gpuva_fn_ops *ops = mgr->ops; + + if (unlikely(!(ops && ops->sm_step_remap && + ops->sm_step_unmap))) + return -EINVAL; + + return __drm_gpuva_sm_unmap(mgr, ops, priv, + req_addr, req_range); +} +EXPORT_SYMBOL (drm_gpuva_sm_unmap); + +static struct drm_gpuva_op * +gpuva_op_alloc (struct drm_gpuva_manager *mgr) +{ + const struct drm_gpuva_fn_ops *fn = mgr->ops; + struct drm_gpuva_op *op; + + if (fn && fn->op_alloc) + op = fn->op_alloc(); + else + op = kzalloc(sizeof(*op), GFP_KERNEL); + + if (unlikely(!op)) + return NULL; + + return op; +} + +static void +gpuva_op_free (struct drm_gpuva_manager *mgr, + struct drm_gpuva_op *op) +{ + const struct drm_gpuva_fn_ops *fn = mgr->ops; + + if (fn && fn->op_free) + fn->op_free(op); + else + kfree(op); +} + +static int +drm_gpuva_sm_step (struct drm_gpuva_op *__op, + void *priv) +{ + struct { + struct drm_gpuva_manager *mgr; + struct drm_gpuva_ops *ops; + } *args = priv; + struct drm_gpuva_manager *mgr = args->mgr; + struct drm_gpuva_ops *ops = args->ops; + struct drm_gpuva_op *op; + + op = gpuva_op_alloc(mgr); + if (unlikely(!op)) + goto err; + + memcpy(op, __op, sizeof(*op)); + + if (op->op == DRM_GPUVA_OP_REMAP) { + struct drm_gpuva_op_remap *__r = &__op->remap; + struct drm_gpuva_op_remap *r = &op->remap; + + r->unmap = kmemdup(__r->unmap, sizeof(*r->unmap), + GFP_KERNEL); + if (unlikely(!r->unmap)) + goto err_free_op; + + if (__r->prev) { + r->prev = kmemdup(__r->prev, sizeof(*r->prev), + GFP_KERNEL); + if (unlikely(!r->prev)) + goto err_free_unmap; + } + + if (__r->next) { + r->next = kmemdup(__r->next, sizeof(*r->next), + GFP_KERNEL); + if (unlikely(!r->next)) + goto err_free_prev; + } + } + + list_add_tail(&op->entry, &ops->list); + + return 0; + +err_free_unmap: + kfree(op->remap.unmap); +err_free_prev: + kfree(op->remap.prev); +err_free_op: + gpuva_op_free(mgr, op); +err: + return -ENOMEM; +} + +static const struct drm_gpuva_fn_ops gpuva_list_ops = { + .sm_step_map = drm_gpuva_sm_step, + .sm_step_remap = drm_gpuva_sm_step, + .sm_step_unmap = drm_gpuva_sm_step, +}; + +/** +* drm_gpuva_sm_map_ops_create - creates the &drm_gpuva_ops to split and merge
+ * @mgr: the&drm_gpuva_managerrepresenting the GPU VA space
+ * @req_addr: the start address of the new mapping
+ * @req_range: the range of the new mapping
+ * @req_obj: the&drm_gem_objectto map
+ * @req_offset: the offset within the&drm_gem_object
+ *
+ * This function creates a list of operations to perform splitting and merging
+ * of existent mapping(s) with the newly requested one.
+ *
+ * The list can be iterated with&drm_gpuva_for_each_opand must be processed
+ * in the given order. It can contain map, unmap and remap operations, but it
+ * also can be empty if no operation is required, e.g. if the requested mapping
+ * already exists is the exact same way.
+ *
+ * There can be an arbitrary amount of unmap operations, a maximum of two remap
+ * operations and a single map operation. The latter one represents the original
+ * map operation requested by the caller.
+ *
+ * Note that before calling this function again with another mapping request it
+ * is necessary to update the &drm_gpuva_manager's view of the GPU VA space. The
+ * previously obtained operations must be either processed or abandoned. To
+ * update the &drm_gpuva_manager's view of the GPU VA spacedrm_gpuva_insert(),
+ *drm_gpuva_destroy_locked()and/ordrm_gpuva_destroy_unlocked()should be
+ * used.
+ *
+ * After the caller finished processing the returned&drm_gpuva_ops, they must
+ * be freed with&drm_gpuva_ops_free.
+ *
+ * Returns: a pointer to the&drm_gpuva_opson success, anERR_PTRon failure
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +37
-0+struct drm_gpuva_ops * +drm_gpuva_sm_map_ops_create (struct drm_gpuva_manager *mgr, + u64 req_addr, u64 req_range, + struct drm_gem_object *req_obj, u64 req_offset) +{ + struct drm_gpuva_ops *ops; + struct { + struct drm_gpuva_manager *mgr; + struct drm_gpuva_ops *ops; + } args; + int ret; + + ops = kzalloc(sizeof(*ops), GFP_KERNEL); + if (unlikely(!ops)) + return ERR_PTR(-ENOMEM); + + INIT_LIST_HEAD(&ops->list); + + args.mgr = mgr; + args.ops = ops; + + ret = __drm_gpuva_sm_map(mgr, &gpuva_list_ops, &args, + req_addr, req_range, + req_obj, req_offset); + if (ret) + goto err_free_ops; + + return ops; + +err_free_ops: + drm_gpuva_ops_free(mgr, ops); + return ERR_PTR(ret); +} +EXPORT_SYMBOL (drm_gpuva_sm_map_ops_create); + +/** +* drm_gpuva_sm_unmap_ops_create - creates the &drm_gpuva_ops to split on unmap
+ * @mgr: the&drm_gpuva_managerrepresenting the GPU VA space
+ * @req_addr: the start address of the range to unmap
+ * @req_range: the range of the mappings to unmap
+ *
+ * This function creates a list of operations to perform unmapping and, if
+ * required, splitting of the mappings overlapping the unmap range.
+ *
+ * The list can be iterated with&drm_gpuva_for_each_opand must be processed
+ * in the given order. It can contain unmap and remap operations, depending on
+ * whether there are actual overlapping mappings to split.
+ *
+ * There can be an arbitrary amount of unmap operations and a maximum of two
+ * remap operations.
+ *
+ * Note that before calling this function again with another range to unmap it
+ * is necessary to update the &drm_gpuva_manager's view of the GPU VA space. The
+ * previously obtained operations must be processed or abandoned. To update the
+ * &drm_gpuva_manager's view of the GPU VA spacedrm_gpuva_insert(),
+ *drm_gpuva_destroy_locked()and/ordrm_gpuva_destroy_unlocked()should be
+ * used.
+ *
+ * After the caller finished processing the returned&drm_gpuva_ops, they must
+ * be freed with&drm_gpuva_ops_free.
+ *
+ * Returns: a pointer to the&drm_gpuva_opson success, anERR_PTRon failure
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +35
-0+struct drm_gpuva_ops * +drm_gpuva_sm_unmap_ops_create (struct drm_gpuva_manager *mgr, + u64 req_addr, u64 req_range) +{ + struct drm_gpuva_ops *ops; + struct { + struct drm_gpuva_manager *mgr; + struct drm_gpuva_ops *ops; + } args; + int ret; + + ops = kzalloc(sizeof(*ops), GFP_KERNEL); + if (unlikely(!ops)) + return ERR_PTR(-ENOMEM); + + INIT_LIST_HEAD(&ops->list); + + args.mgr = mgr; + args.ops = ops; + + ret = __drm_gpuva_sm_unmap(mgr, &gpuva_list_ops, &args, + req_addr, req_range); + if (ret) + goto err_free_ops; + + return ops; + +err_free_ops: + drm_gpuva_ops_free(mgr, ops); + return ERR_PTR(ret); +} +EXPORT_SYMBOL (drm_gpuva_sm_unmap_ops_create); + +/** +* drm_gpuva_prefetch_ops_create - creates the &drm_gpuva_ops to prefetch
+ * @mgr: the&drm_gpuva_managerrepresenting the GPU VA space
+ * @addr: the start address of the range to prefetch
+ * @range: the range of the mappings to prefetch
+ *
+ * This function creates a list of operations to perform prefetching.
+ *
+ * The list can be iterated with&drm_gpuva_for_each_opand must be processed
+ * in the given order. It can contain prefetch operations.
+ *
+ * There can be an arbitrary amount of prefetch operations.
+ *
+ * After the caller finished processing the returned&drm_gpuva_ops, they must
+ * be freed with&drm_gpuva_ops_free.
+ *
+ * Returns: a pointer to the&drm_gpuva_opson success, anERR_PTRon failure
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +38
-0+struct drm_gpuva_ops * +drm_gpuva_prefetch_ops_create (struct drm_gpuva_manager *mgr, + u64 addr, u64 range) +{ + struct drm_gpuva_ops *ops; + struct drm_gpuva_op *op; + struct drm_gpuva *va; + u64 end = addr + range; + int ret; + + ops = kzalloc(sizeof(*ops), GFP_KERNEL); + if (!ops) + return ERR_PTR(-ENOMEM); + + INIT_LIST_HEAD(&ops->list); + + drm_gpuva_for_each_va_range(va, mgr, addr, end) { + op = gpuva_op_alloc(mgr); + if (!op) { + ret = -ENOMEM; + goto err_free_ops; + } + + op->op = DRM_GPUVA_OP_PREFETCH; + op->prefetch.va = va; + list_add_tail(&op->entry, &ops->list); + } + + return ops; + +err_free_ops: + drm_gpuva_ops_free(mgr, ops); + return ERR_PTR(ret); +} +EXPORT_SYMBOL (drm_gpuva_prefetch_ops_create); + +/** +* drm_gpuva_gem_unmap_ops_create - creates the &drm_gpuva_ops to unmap a GEM
+ * @mgr: the&drm_gpuva_managerrepresenting the GPU VA space
+ * @obj: the&drm_gem_objectto unmap
+ *
+ * This function creates a list of operations to perform unmapping for every
+ * GPUVA attached to a GEM.
+ *
+ * The list can be iterated with&drm_gpuva_for_each_opand consists out of an
+ * arbitrary amount of unmap operations.
+ *
+ * After the caller finished processing the returned&drm_gpuva_ops, they must
+ * be freed with&drm_gpuva_ops_free.
+ *
+ * It is the callers responsibility to protect the GEMs GPUVA list against
+ * concurrent access using the GEMsdma_resvlock.
+ *
+ * Returns: a pointer to the&drm_gpuva_opson success, anERR_PTRon failure
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +47
-0+struct drm_gpuva_ops * +drm_gpuva_gem_unmap_ops_create (struct drm_gpuva_manager *mgr, + struct drm_gem_object *obj) +{ + struct drm_gpuva_ops *ops; + struct drm_gpuva_op *op; + struct drm_gpuva *va; + int ret; + + if (drm_gpuva_manager_external_lock(mgr)) + drm_gpuva_manager_ext_assert_held(mgr); + else + dma_resv_assert_held(obj->resv); + + ops = kzalloc(sizeof(*ops), GFP_KERNEL); + if (!ops) + return ERR_PTR(-ENOMEM); + + INIT_LIST_HEAD(&ops->list); + + drm_gem_for_each_gpuva(va, obj) { + op = gpuva_op_alloc(mgr); + if (!op) { + ret = -ENOMEM; + goto err_free_ops; + } + + op->op = DRM_GPUVA_OP_UNMAP; + op->unmap.va = va; + list_add_tail(&op->entry, &ops->list); + } + + return ops; + +err_free_ops: + drm_gpuva_ops_free(mgr, ops); + return ERR_PTR(ret); +} +EXPORT_SYMBOL (drm_gpuva_gem_unmap_ops_create); + + +/** +* drm_gpuva_ops_free - free the given &drm_gpuva_ops +* @mgr: the &drm_gpuva_manager the ops were created for +* @ops: the &drm_gpuva_ops to free +* +* Frees the given &drm_gpuva_ops structure including all the ops associated
+ * with it.
+ */drivers/gpu/drm/drm_gpuva_mgr.c new file +21
-0+void +drm_gpuva_ops_free (struct drm_gpuva_manager *mgr, + struct drm_gpuva_ops *ops) +{ + struct drm_gpuva_op *op, *next; + + drm_gpuva_for_each_op_safe(op, next, ops) { + list_del(&op->entry); + + if (op->op == DRM_GPUVA_OP_REMAP) { + kfree(op->remap.prev); + kfree(op->remap.next); + kfree(op->remap.unmap); + } + + gpuva_op_free(mgr, op); + } + + kfree(ops); +} +EXPORT_SYMBOL (drm_gpuva_ops_free);include/drm/drm_drv.h +6
-0@@ -104,6 +104,12 @@ enum drm_driver_feature {* acceleration should be handled by two drivers that are connected using auxiliary bus. */ DRIVER_COMPUTE_ACCEL = BIT(7), + /** + * @DRIVER_GEM_GPUVA: + * + * Driver supports user defined GPU VA bindings for GEM objects. + */ + DRIVER_GEM_GPUVA = BIT(8), /* IMPORTANT: Below are all the legacy flags, add new ones above. */include/drm/drm_gem.h +21
-0@@ -36,6 +36,8 @@ #include <linux/kref.h> #include <linux/dma-resv.h> +#include <linux/list.h> +#include <linux/mutex.h> #include <drm/drm_vma_manager.h>@@ -379,6 +381,18 @@ struct drm_gem_object {*/ struct dma_resv _resv; + /** + * @gpuva: + * + * Provides the list of GPU VAs attached to this GEM object. + * + * Drivers should lock list accesses with the GEMs &dma_resv lock + * (&drm_gem_object.resv). + */ + struct { + struct list_head list; + } gpuva; + /*** @funcs: * @@ -526,4 +540,42 @@ unsigned long drm_gem_lru_scan(structdrm_gem_lru *lru, int drm_gem_evict(struct drm_gem_object *obj); +/** +* drm_gem_gpuva_init - initialize the gpuva list of a GEM object +* @obj: the &drm_gem_object +* +* This initializes the &drm_gem_object's &drm_gpuva list. +* +* Calling this function is only necessary for drivers intending to support the
+ *&drm_driver_featureDRIVER_GEM_GPUVA.
+ */include/drm/drm_gem.h unsigned long drm_gem_lru_scan(struct drm_gem_lru *lru, · line 558
Quoted 7 lines
include/drm/drm_gem.h +7
-0+static inline void drm_gem_gpuva_init (struct drm_gem_object *obj) +{ + INIT_LIST_HEAD(&obj->gpuva.list); +} + +/** +* drm_gem_for_each_gpuva - iternator to walk over a list of gpuvas s/iternator/iterator/. (multiple places) Although since "iterator" typically refers to an object being iterated over, perhaps
"
drm_gem_for_each_gpuva()- iterate over a list of gpuvas", (general comment across the patch).Quoted 871 lines
+ * @entry:
&drm_gpuvastructure to assign to in each iteration step
+ * @obj: the&drm_gem_objectthe&drm_gpuvasto walk are associated with
+ *
+ * This iterator walks over all&drm_gpuvastructures associated with the
+ *&drm_gpuva_manager.
+ */include/drm/drm_gem.h +5
-0+#define drm_gem_for_each_gpuva (entry__, obj__) \ + list_for_each_entry(entry__, &(obj__)->gpuva.list, gem.entry) + +/** +* drm_gem_for_each_gpuva_safe - iternator to safely walk over a list of gpuvas
+ * @entry:&drm_gpuvastructure to assign to in each iteration step
+ * @next: &next&drm_gpuvato store the next step
+ * @obj: the&drm_gem_objectthe&drm_gpuvasto walk are associated with
+ *
+ * This iterator walks over all&drm_gpuvastructures associated with the
+ *&drm_gem_object. It is implemented withlist_for_each_entry_safe(), hence
+ * it is save against removal of elements.
+ */
+#definedrm_gem_for_each_gpuva_safe(entry__, next__, obj__)\
+list_for_each_entry_safe(entry__, next__, &(obj__)->gpuva.list, gem.entry)
+
#endif /*__DRM_GEM_H__*/include/drm/drm_gpuva_mgr.h new file +9
-0@@ -0,0 +1,756 @@ +/* SPDX-License-Identifier: GPL-2.0 */ + +#ifndef __DRM_GPUVA_MGR_H__ +#define __DRM_GPUVA_MGR_H__ + +/* +* Copyright (c) 2022 Red Hat. +* +* Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */include/drm/drm_gpuva_mgr.h new file +18
-0+ +#include <linux/list.h> +#include <linux/rbtree.h> +#include <linux/types.h> + +#include <drm/drm_gem.h> + +struct drm_gpuva_manager ; +struct drm_gpuva_fn_ops ; + +/** +* enum drm_gpuva_flags - flags for struct drm_gpuva +*/ +enum drm_gpuva_flags { + /** +* @DRM_GPUVA_INVALIDATED: +* +* Flag indicating that the &drm_gpuva's backing GEM is invalidated.
+ */
+DRM_GPUVA_INVALIDATED= (1 << 0),include/drm/drm_gpuva_mgr.h new file +35
-0+ + /** + * @DRM_GPUVA_SPARSE: + * + * Flag indicating that the &drm_gpuva is a sparse mapping. + */ + DRM_GPUVA_SPARSE = (1 << 1), + + /** + * @DRM_GPUVA_USERBITS: user defined bits + */ + DRM_GPUVA_USERBITS = (1 << 2), +}; + +/** +* struct drm_gpuva - structure to track a GPU VA mapping +* +* This structure represents a GPU VA mapping and is associated with a +* &drm_gpuva_manager. +* +* Typically, this structure is embedded in bigger driver structures. +*/ +struct drm_gpuva { + /** + * @mgr: the &drm_gpuva_manager this object is associated with + */ + struct drm_gpuva_manager *mgr; + + /** + * @flags: the &drm_gpuva_flags for this mapping + */ + enum drm_gpuva_flags flags; + + /** + * @va: structure containing the address and range of the
&drm_gpuva
+ */
+ struct {
+ /**
+ * @addr: the start address
+ */
+ u64 addr;include/drm/drm_gpuva_mgr.h new file +23
-0+ + /* + * @range: the range + */ + u64 range; + } va; + + /** + * @gem: structure containing the &drm_gem_object and it's offset + */ + struct { + /** + * @offset: the offset within the &drm_gem_object + */ + u64 offset; + + /** + * @obj: the mapped &drm_gem_object + */ + struct drm_gem_object *obj; + + /** + * @entry: the &list_head to attach this object to a
&drm_gem_object
+ */
+ structlist_headentry;
+ } gem;
+
+ /**
+ * @rb: structure containing data to store&drm_gpuvasin a rb-tree
+ */
+ struct {
+ /**
+ * @rb: the rb-tree node
+ */
+ structrb_nodenode;
+
+ /**
+ * @entry: The&list_headto additionally connect&drm_gpuvas+ * in the same order they appear in the interval tree. This is + * useful to keep iterating&drm_gpuvasfrom a start node found
+ * through the rb-tree while doing modifications on the rb-tree
+ * itself.
+ */
+ structlist_headentry;
+
+ /**
+ * @__subtree_last: needed by the interval tree, holding last-in-subtree
+ */
+ u64__subtree_last;
+ } rb;include/drm/drm_gpuva_mgr.h new file +3
-0+}; + +int drm_gpuva_insert(struct drm_gpuva_manager *mgr, struct
drm_gpuva*va);include/drm/drm_gpuva_mgr.h new file +10
-0+void drm_gpuva_remove (struct drm_gpuva *va); + +void drm_gpuva_link (struct drm_gpuva *va); +void drm_gpuva_unlink (struct drm_gpuva *va); + +struct drm_gpuva *drm_gpuva_find (struct drm_gpuva_manager *mgr, + u64 addr, u64 range); +struct drm_gpuva *drm_gpuva_find_first (struct drm_gpuva_manager *mgr, + u64 addr, u64 range); +struct drm_gpuva *drm_gpuva_find_prev(struct drm_gpuva_manager*mgr, u64 start);
+structdrm_gpuva*drm_gpuva_find_next(structdrm_gpuva_manager*mgr, u64 end);include/drm/drm_gpuva_mgr.h new file +2
-0+ +bool drm_gpuva_interval_empty(struct drm_gpuva_manager *mgr, u64addr, u64 range);
include/drm/drm_gpuva_mgr.h new file +2
-0+ +static inline void drm_gpuva_init(struct drm_gpuva *va, u64 addr,u64 range,
+ structdrm_gem_object*obj, u64 offset)include/drm/drm_gpuva_mgr.h new file +9
-0+{ + va->va.addr = addr; + va->va.range = range; + va->gem.obj = obj; + va->gem.offset = offset; +} + +/** +* drm_gpuva_invalidate - sets whether the backing GEM of this
&drm_gpuvais
+ * invalidated
+ * @va: the&drm_gpuvato set the invalidate flag for
+ * @invalidate: indicates whether the&drm_gpuvais invalidated
+ */
+static inline voiddrm_gpuva_invalidate(struct drm_gpuva *va, bool invalidate)include/drm/drm_gpuva_mgr.h new file +9
-0+{ + if (invalidate) + va->flags |= DRM_GPUVA_INVALIDATED; + else + va->flags &= ~DRM_GPUVA_INVALIDATED; +} + +/** +* drm_gpuva_invalidated - indicates whether the backing BO of this
&drm_gpuva
+ * is invalidated
+ * @va: the&drm_gpuvato check
+ */include/drm/drm_gpuva_mgr.h new file +9
-0+static inline bool drm_gpuva_invalidated (struct drm_gpuva *va) +{ + return va->flags & DRM_GPUVA_INVALIDATED; +} + +#ifdef CONFIG_LOCKDEP +typedef struct lockdep_map *lockdep_map_p ; +#define drm_gpuva_manager_ext_assert_held (mgr) \ + lockdep_assert(lock_is_held((mgr)->ext_lock) !=
LOCK_STATE_NOT_HELD)
+/**
+ *drm_gpuva_manager_set_ext_lock- set the external lock according to
+ * @DRM_GPUVA_MANAGER_LOCK_EXTERN
+ * @mgr: the&drm_gpuva_managerto set the lock for
+ * @lock: the lock to set
+ *
+ * If @DRM_GPUVA_MANAGER_LOCK_EXTERN is set, drivers need to call this function
+ * to provide the lock used to lock linking and unlinking of&drm_gpuvasto the
+ *&drm_gem_objectsGPUVA list.
+ */include/drm/drm_gpuva_mgr.h new file +5
-0+#define drm_gpuva_manager_set_ext_lock (mgr, lock) \ + (mgr)->ext_lock = &(lock)->dep_map +#else +typedef struct {/* nothing */ }lockdep_map_p ; +#define drm_gpuva_manager_ext_assert_held (mgr) do {(void)(mgr); } while (0)
include/drm/drm_gpuva_mgr.h new file +5
-0+#define drm_gpuva_manager_set_ext_lock (mgr, lock) do { } while (0) +#endif + +/** +* enum drm_gpuva_manager_flags - the feature flags for the
&drm_gpuva_manager
+ */include/drm/drm_gpuva_mgr.h new file +17
-0+enum drm_gpuva_manager_flags { + /** + * @DRM_GPUVA_MANAGER_LOCK_EXTERN: + * + * Indicates the driver has it's own external lock for linking and + * unlinking &drm_gpuvas to the &drm_gem_objects GPUVA list. + * + * When setting this flag it is rquired to set a lock via + * drm_gpuva_set_ext_lock(). + */ + DRM_GPUVA_MANAGER_LOCK_EXTERN = (1 << 0), +}; + +/** +* struct drm_gpuva_manager - DRM GPU VA Manager +* +* The DRM GPU VA Manager keeps track of a GPU's virtual address space by using
+ *&maple_treestructures. Typically, this structure is embedded in bigger
+ * driver structures.
+ *
+ * Drivers can pass addresses and ranges in an arbitrary unit, e.g. bytes or
+ * pages.
+ *
+ * There should be one manager instance per GPU virtual address space.
+ */include/drm/drm_gpuva_mgr.h new file +46
-0+struct drm_gpuva_manager { + /** + * @name: the name of the DRM GPU VA space + */ + const char *name; + + /** + * @flags: the feature flags of the &drm_gpuva_manager + */ + enum drm_gpuva_manager_flags flags; + + /** + * @mm_start: start of the VA space + */ + u64 mm_start; + + /** + * @mm_range: length of the VA space + */ + u64 mm_range; + + /** + * @rb: structures to track &drm_gpuva entries + */ + struct { + /** + * @tree: the rb-tree to track GPU VA mappings + */ + struct rb_root_cached tree; + + /** + * @list: the &list_head to track GPU VA mappings + */ + struct list_head list; + } rb; + + /** + * @kernel_alloc_node: + * + * &drm_gpuva representing the address space cutout reserved for + * the kernel + */ + struct drm_gpuva kernel_alloc_node; + + /** + * @ops: &drm_gpuva_fn_ops providing the split/merge steps todrivers
+ */
+ const structdrm_gpuva_fn_ops*ops;
+
+ /**
+ * @ext_lock:&lockdep_mapaccording to
@DRM_GPUVA_MANAGER_LOCK_EXTERN
+ */
+lockdep_map_pext_lock;include/drm/drm_gpuva_mgr.h new file +16
-0+}; + +void drm_gpuva_manager_init (struct drm_gpuva_manager *mgr, + const char *name, + u64 start_offset, u64 range, + u64 reserve_offset, u64 reserve_range, + const struct drm_gpuva_fn_ops *ops, + enum drm_gpuva_manager_flags flags); +void drm_gpuva_manager_destroy (struct drm_gpuva_manager *mgr); + +/** +* drm_gpuva_manager_external_lock - indicates whether the +* @DRM_GPUVA_MANAGER_LOCK_EXTERN flag is set +* @mgr: the &drm_gpuva_manager to check the flag for +*/ +static inline bool drm_gpuva_manager_external_lock(struct
drm_gpuva_manager*mgr)include/drm/drm_gpuva_mgr.h new file +6
-0+{ + return mgr->flags & DRM_GPUVA_MANAGER_LOCK_EXTERN; +} + +/** +* drm_gpuva_for_each_va_range - iternator to walk over a range of
&drm_gpuvas
+ * @va__:&drm_gpuvastructure to assign to in each iteration step
+ * @mgr__:&drm_gpuva_managerto walk over
+ * @start__: starting offset, the first gpuva will overlap this
+ * @end__: ending offset, the last gpuva will start before this (but may
+ * overlap)
+ *
+ * This iterator walks over all&drm_gpuvasin the&drm_gpuva_managerthat lie
+ * between @start__ and @end__. It is implemented similarly tolist_for_each(),
+ * but is using the &drm_gpuva_manager's internal interval tree to accelerate
+ * the search for the starting&drm_gpuva, and hence isn't safe against removal
+ * of elements. It assumes that @end__ is within (or is the upper limit of) the
+ *&drm_gpuva_manager. This iterator does not skip over the &drm_gpuva_manager's
+ * @kernel_alloc_node.
+ */include/drm/drm_gpuva_mgr.h new file +8
-0+#define drm_gpuva_for_each_va_range (va__, mgr__, start__, end__) \ + for (va__ = drm_gpuva_find_first((mgr__), (start__), (end__)); \ + va__ && (va__->va.addr < (end__)) && \ + !list_entry_is_head(va__, &(mgr__)->rb.list, rb.entry); \ + va__ = list_next_entry(va__, rb.entry)) + +/** +* drm_gpuva_for_each_va_range_safe - iternator to safely walk over a range of
+ *&drm_gpuvas
+ * @va__:&drm_gpuvato assign to in each iteration step
+ * @next__: another&drm_gpuvato use as temporary storage
+ * @mgr__:&drm_gpuva_managerto walk over
+ * @start__: starting offset, the first gpuva will overlap this
+ * @end__: ending offset, the last gpuva will start before this (but may
+ * overlap)
+ *
+ * This iterator walks over all&drm_gpuvasin the&drm_gpuva_managerthat lie
+ * between @start__ and @end__. It is implemented similarly to
+ *list_for_each_safe(), but is using the &drm_gpuva_manager's internal interval
+ * tree to accelerate the search for the starting&drm_gpuva, and hence is safe
+ * against removal of elements. It assumes that @end__ is within (or is the
+ * upper limit of) the&drm_gpuva_manager. This iterator does not skip over the
+ * &drm_gpuva_manager's @kernel_alloc_node.
+ */
+#definedrm_gpuva_for_each_va_range_safe(va__, next__, mgr__, start__, end__)\
+ for (va__=drm_gpuva_find_first((mgr__), (start__), (end__)), \
+next__= va ?list_next_entry(va__, rb.entry): NULL; \
+va__&& (va__->va.addr< (end__)) && \
+ !list_entry_is_head(va__, &(mgr__)->rb.list, rb.entry); \
+va__=next__,next__=list_next_entry(va__, rb.entry))include/drm/drm_gpuva_mgr.h new file +7
-0+ +/** +* drm_gpuva_for_each_va - iternator to walk over all &drm_gpuvas +* @va__: &drm_gpuva to assign to in each iteration step +* @mgr__: &drm_gpuva_manager to walk over +* +* This iterator walks over all &drm_gpuva structures associated with the given
+ *&drm_gpuva_manager.
+ */include/drm/drm_gpuva_mgr.h new file +5
-0+#define drm_gpuva_for_each_va (va__, mgr__) \ + list_for_each_entry(va__, &(mgr__)->rb.list, rb.entry) + +/** +* drm_gpuva_for_each_va_safe - iternator to safely walk over all
&drm_gpuvas
+ * @va__:&drm_gpuvato assign to in each iteration step
+ * @next__: another&drm_gpuvato use as temporary storage
+ * @mgr__:&drm_gpuva_managerto walk over
+ *
+ * This iterator walks over all&drm_gpuvastructures associated with the given
+ *&drm_gpuva_manager. It is implemented withlist_for_each_entry_safe(), and
+ * hence safe against the removal of elements.
+ */
+#definedrm_gpuva_for_each_va_safe(va__, next__, mgr__)\
+list_for_each_entry_safe(va__, next__, &(mgr__)->rb.list, rb.entry)include/drm/drm_gpuva_mgr.h new file +5
-0+ +/** +* enum drm_gpuva_op_type - GPU VA operation type +* +* Operations to alter the GPU VA mappings tracked by the
&drm_gpuva_manager.
+ */include/drm/drm_gpuva_mgr.h new file +77
-0+enum drm_gpuva_op_type { + /** + * @DRM_GPUVA_OP_MAP: the map op type + */ + DRM_GPUVA_OP_MAP, + + /** + * @DRM_GPUVA_OP_REMAP: the remap op type + */ + DRM_GPUVA_OP_REMAP, + + /** + * @DRM_GPUVA_OP_UNMAP: the unmap op type + */ + DRM_GPUVA_OP_UNMAP, + + /** + * @DRM_GPUVA_OP_PREFETCH: the prefetch op type + */ + DRM_GPUVA_OP_PREFETCH, +}; + +/** +* struct drm_gpuva_op_map - GPU VA map operation +* +* This structure represents a single map operation generated by the +* DRM GPU VA manager. +*/ +struct drm_gpuva_op_map { + /** + * @va: structure containing address and range of a map + * operation + */ + struct { + /** + * @addr: the base address of the new mapping + */ + u64 addr; + + /** + * @range: the range of the new mapping + */ + u64 range; + } va; + + /** + * @gem: structure containing the &drm_gem_object and it's offset + */ + struct { + /** + * @offset: the offset within the &drm_gem_object + */ + u64 offset; + + /** + * @obj: the &drm_gem_object to map + */ + struct drm_gem_object *obj; + } gem; +}; + +/** +* struct drm_gpuva_op_unmap - GPU VA unmap operation +* +* This structure represents a single unmap operation generated by the +* DRM GPU VA manager. +*/ +struct drm_gpuva_op_unmap { + /** + * @va: the &drm_gpuva to unmap + */ + struct drm_gpuva *va; + + /** +* @keep: +* +* Indicates whether this &drm_gpuva is physically contiguous with the
+ * original mapping request.
+ *
+ * Optionally, if &keep is set, drivers may keep the actual page table
+ * mappings for this&drm_gpuva, adding the missing page table entries
+ * only and update the&drm_gpuva_manageraccordingly. + */
+ bool keep;include/drm/drm_gpuva_mgr.h new file +6
-0+}; + +/** +* struct drm_gpuva_op_remap - GPU VA remap operation +* +* This represents a single remap operation generated by the DRM GPU VA manager.
+ *
+ * A remap operation is generated when an existing GPU VA mmapping is split up
+ * by inserting a new GPU VA mapping or by partially unmapping existent
+ * mapping(s), hence it consists of a maximum of two map and one unmap
+ * operation.
+ *
+ * The @unmap operation takes care of removing the original existing mapping.
+ * @prev is used to remap the preceding part, @next the subsequent part.
+ *
+ * If either a new mapping's start address is aligned with the start address
+ * of the old mapping or the new mapping's end address is aligned with the
+ * end address of the old mapping, either @prev or @next is NULL.
+ *
+ * Note, the reason for a dedicated remap operation, rather than arbitrary
+ * unmap and map operations, is to give drivers the chance of extracting driver
+ * specific data for creating the new mappings from the unmap operations's
+ *&drm_gpuvastructure which typically is embedded in larger driver specific
+ * structures.
+ */include/drm/drm_gpuva_mgr.h new file +21
-0+struct drm_gpuva_op_remap { + /** + * @prev: the preceding part of a split mapping + */ + struct drm_gpuva_op_map *prev; + + /** + * @next: the subsequent part of a split mapping + */ + struct drm_gpuva_op_map *next; + + /** + * @unmap: the unmap operation for the original existing mapping + */ + struct drm_gpuva_op_unmap *unmap; +}; + +/** +* struct drm_gpuva_op_prefetch - GPU VA prefetch operation +* +* This structure represents a single prefetch operation generated by the
+ * DRM GPU VA manager.
+ */include/drm/drm_gpuva_mgr.h new file +19
-0+struct drm_gpuva_op_prefetch { + /** + * @va: the &drm_gpuva to prefetch + */ + struct drm_gpuva *va; +}; + +/** +* struct drm_gpuva_op - GPU VA operation +* +* This structure represents a single generic operation. +* +* The particular type of the operation is defined by @op. +*/ +struct drm_gpuva_op { + /** +* @entry: +* +* The &list_head used to distribute instances of this struct within
+ *&drm_gpuva_ops.
+ */
+ structlist_headentry;include/drm/drm_gpuva_mgr.h new file +47
-0+ + /** + * @op: the type of the operation + */ + enum drm_gpuva_op_type op; + + union { + /** + * @map: the map operation + */ + struct drm_gpuva_op_map map; + + /** + * @remap: the remap operation + */ + struct drm_gpuva_op_remap remap; + + /** + * @unmap: the unmap operation + */ + struct drm_gpuva_op_unmap unmap; + + /** + * @prefetch: the prefetch operation + */ + struct drm_gpuva_op_prefetch prefetch; + }; +}; + +/** +* struct drm_gpuva_ops - wraps a list of &drm_gpuva_op +*/ +struct drm_gpuva_ops { + /** + * @list: the &list_head + */ + struct list_head list; +}; + +/** +* drm_gpuva_for_each_op - iterator to walk over &drm_gpuva_ops +* @op: &drm_gpuva_op to assign in each iteration step +* @ops: &drm_gpuva_ops to walk +* +* This iterator walks over all ops within a given list of operations. +*/ +#define drm_gpuva_for_each_op (op, ops) list_for_each_entry(op,&(ops)->list, entry)
include/drm/drm_gpuva_mgr.h new file +3
-0+ +/** +* drm_gpuva_for_each_op_safe - iterator to safely walk over
&drm_gpuva_ops
+ * @op:&drm_gpuva_opto assign in each iteration step
+ * @next: &next&drm_gpuva_opto store the next step
+ * @ops:&drm_gpuva_opsto walk
+ *
+ * This iterator walks over all ops within a given list of operations. It is
+ * implemented withlist_for_each_safe(), so save against removal of elements.
+ */include/drm/drm_gpuva_mgr.h new file +5
-0+#define drm_gpuva_for_each_op_safe (op, next, ops) \ + list_for_each_entry_safe(op, next, &(ops)->list, entry) + +/** +* drm_gpuva_for_each_op_from_reverse - iterate backwards from the given point
+ * @op:&drm_gpuva_opto assign in each iteration step
+ * @ops:&drm_gpuva_opsto walk
+ *
+ * This iterator walks over all ops within a given list of operations beginning
+ * from the given operation in reverse order.
+ */include/drm/drm_gpuva_mgr.h new file +5
-0+#define drm_gpuva_for_each_op_from_reverse (op, ops) \ + list_for_each_entry_from_reverse(op, &(ops)->list, entry) + +/** +* drm_gpuva_first_op - returns the first &drm_gpuva_op from
&drm_gpuva_ops
+ * @ops: the&drm_gpuva_opsto get the fist&drm_gpuva_opfrom
+ */include/drm/drm_gpuva_mgr.h new file +5
-0+#define drm_gpuva_first_op (ops) \ + list_first_entry(&(ops)->list, struct drm_gpuva_op, entry) + +/** +* drm_gpuva_last_op - returns the last &drm_gpuva_op from
&drm_gpuva_ops
+ * @ops: the&drm_gpuva_opsto get the last&drm_gpuva_opfrom
+ */include/drm/drm_gpuva_mgr.h new file +46
-0+#define drm_gpuva_last_op (ops) \ + list_last_entry(&(ops)->list, struct drm_gpuva_op, entry) + +/** +* drm_gpuva_prev_op - previous &drm_gpuva_op in the list +* @op: the current &drm_gpuva_op +*/ +#define drm_gpuva_prev_op (op) list_prev_entry(op, entry) + +/** +* drm_gpuva_next_op - next &drm_gpuva_op in the list +* @op: the current &drm_gpuva_op +*/ +#define drm_gpuva_next_op (op) list_next_entry(op, entry) + +struct drm_gpuva_ops * +drm_gpuva_sm_map_ops_create (struct drm_gpuva_manager *mgr, + u64 addr, u64 range, + struct drm_gem_object *obj, u64 offset); +struct drm_gpuva_ops * +drm_gpuva_sm_unmap_ops_create (struct drm_gpuva_manager *mgr, + u64 addr, u64 range); + +struct drm_gpuva_ops * +drm_gpuva_prefetch_ops_create (struct drm_gpuva_manager *mgr, + u64 addr, u64 range); + +struct drm_gpuva_ops * +drm_gpuva_gem_unmap_ops_create (struct drm_gpuva_manager *mgr, + struct drm_gem_object *obj); + +void drm_gpuva_ops_free (struct drm_gpuva_manager *mgr, + struct drm_gpuva_ops *ops); + +static inline void drm_gpuva_init_from_op (struct drm_gpuva *va, + struct drm_gpuva_op_map *op) +{ + drm_gpuva_init(va, op->va.addr, op->va.range, + op->gem.obj, op->gem.offset); +} + +/** +* struct drm_gpuva_fn_ops - callbacks for split/merge steps +* +* This structure defines the callbacks used by &drm_gpuva_sm_map and +* &drm_gpuva_sm_unmap to provide the split/merge steps for map and unmap
+ * operations to drivers.
+ */include/drm/drm_gpuva_mgr.h new file +27
-0+struct drm_gpuva_fn_ops { + /** + * @op_alloc: called when the &drm_gpuva_manager allocates + * a struct drm_gpuva_op + * + * Some drivers may want to embed struct drm_gpuva_op into driver + * specific structures. By implementing this callback drivers can + * allocate memory accordingly. + * + * This callback is optional. + */ + struct drm_gpuva_op *(*op_alloc)(void); + + /** + * @op_free: called when the &drm_gpuva_manager frees a + * struct drm_gpuva_op + * + * Some drivers may want to embed struct drm_gpuva_op into driver + * specific structures. By implementing this callback drivers can + * free the previously allocated memory accordingly. + * + * This callback is optional. + */ + void (*op_free)(struct drm_gpuva_op *op); + + /** + * @sm_step_map: called from &drm_gpuva_sm_map to finallyinsert the
+ * mapping once all previous steps were completed
+ *
+ * The &priv pointer matches the one the driver passed to + *&drm_gpuva_sm_mapor&drm_gpuva_sm_unmap, respectively. + *
+ * Can be NULL if&drm_gpuva_sm_mapis used.
+ */
+ int (*sm_step_map)(structdrm_gpuva_op*op, void *priv); +
+ /**
+ * @sm_step_remap: called from&drm_gpuva_sm_mapand + *&drm_gpuva_sm_unmapto split up an existent mapping + *
+ * This callback is called when existent mapping needs to be split up.
+ * This is the case when either a newly requested mapping overlaps or
+ * is enclosed by an existent mapping or a partial unmap of an existent
+ * mapping is requested.
+ *
+ * The &priv pointer matches the one the driver passed to + *&drm_gpuva_sm_mapor&drm_gpuva_sm_unmap, respectively. + *
+ * Can be NULL if neither&drm_gpuva_sm_mapnor&drm_gpuva_sm_unmapis
+ * used.
+ */
+ int (*sm_step_remap)(structdrm_gpuva_op*op, void *priv); +
+ /**
+ * @sm_step_unmap: called from&drm_gpuva_sm_mapand + *&drm_gpuva_sm_unmapto unmap an existent mapping + *
+ * This callback is called when existent mapping needs to be unmapped.
+ * This is the case when either a newly requested mapping encloses an
+ * existent mapping or an unmap of an existent mapping is requested.
+ *
+ * The &priv pointer matches the one the driver passed to + *&drm_gpuva_sm_mapor&drm_gpuva_sm_unmap, respectively. + *
+ * Can be NULL if neither&drm_gpuva_sm_mapnor&drm_gpuva_sm_unmapis
+ * used.
+ */
+ int (*sm_step_unmap)(structdrm_gpuva_op*op, void *priv);include/drm/drm_gpuva_mgr.h new file +2
-0+}; +include/drm/drm_gpuva_mgr.h line 750
Quoted 10 lines
include/drm/drm_gpuva_mgr.h new file +10
-0+int drm_gpuva_sm_map (struct drm_gpuva_manager *mgr,void *priv, + u64 addr, u64 range, + struct drm_gem_object *obj, u64 offset); + +int drm_gpuva_sm_unmap (struct drm_gpuva_manager *mgr,void *priv, + u64 addr, u64 range); + +void drm_gpuva_map (struct drm_gpuva_manager *mgr, + struct drm_gpuva *va, + struct drm_gpuva_op_map *op);Missing newline
Quoted 3 lines
+void drm_gpuva_remap(struct
drm_gpuva*prev,
+ structdrm_gpuva*next,
+ structdrm_gpuva_op_remap*op);Missing newline
Quoted 3 lines
include/drm/drm_gpuva_mgr.h new file +3
-0+void drm_gpuva_unmap (struct drm_gpuva_op_unmap *op); + +#endif /* __DRM_GPUVA_MGR_H__ */
Re: [PATCH drm-next v6 02/13] drm: manager to keep track of GPUs VA mappings
In reply to Danilo KrummrichQuoted 9 lines · Danilo Krummrich
include/drm/drm_gpuva_mgr.h new file +8
-0+#ifdef CONFIG_LOCKDEP +typedef struct lockdep_map *lockdep_map_p ; +#define drm_gpuva_manager_ext_assert_held (mgr) \ + lockdep_assert(lock_is_held((mgr)->ext_lock) != LOCK_STATE_NOT_HELD) +/** +* drm_gpuva_manager_set_ext_lock - set the external lock according to +* @DRM_GPUVA_MANAGER_LOCK_EXTERN +* @mgr: the &drm_gpuva_manager to set the lock for
include/drm/drm_gpuva_mgr.h line 205
Quoted 10 lines
include/drm/drm_gpuva_mgr.h new file +10
-0+* @lock: the lock to set +* +* If @DRM_GPUVA_MANAGER_LOCK_EXTERN is set, drivers need to call this function +* to provide the lock used to lock linking and unlinking of &drm_gpuvas to the +* &drm_gem_objects GPUVA list. +*/ +#define drm_gpuva_manager_set_ext_lock (mgr, lock) \ + (mgr)->ext_lock = &(lock)->dep_map +#else +typedef struct {/* nothing */ }lockdep_map_p ;
lockdep_map_p conflicts with an identical typedef in maple_tree.h when CONFIG_LOCKDEP is not set (it's being pulled in by mm.h in drm_vma_manager.h). I'll just comment the line out for now.
Quoted 4 lines
include/drm/drm_gpuva_mgr.h new file +4
-0+#define drm_gpuva_manager_ext_assert_held (mgr) do { (void)(mgr); } while (0) +#define drm_gpuva_manager_set_ext_lock (mgr, lock) do { } while (0) +#endif +
Re: [PATCH drm-next v6 02/13] drm: manager to keep track of GPUs VA mappings
Hi Donald,
Quoted 16 lines · Donald Robson
Quoted 15 lines · Danilo Krummrich
include/drm/drm_gpuva_mgr.h new file +14
-0+#ifdef CONFIG_LOCKDEP +typedef struct lockdep_map *lockdep_map_p ; +#define drm_gpuva_manager_ext_assert_held (mgr) \ + lockdep_assert(lock_is_held((mgr)->ext_lock) != LOCK_STATE_NOT_HELD) +/** +* drm_gpuva_manager_set_ext_lock - set the external lock according to +* @DRM_GPUVA_MANAGER_LOCK_EXTERN +* @mgr: the &drm_gpuva_manager to set the lock for +* @lock: the lock to set +* +* If @DRM_GPUVA_MANAGER_LOCK_EXTERN is set, drivers need to call this function +* to provide the lock used to lock linking and unlinking of &drm_gpuvas to the +* &drm_gem_objects GPUVA list. +*/
Quoted 8 lines
include/drm/drm_gpuva_mgr.h line 205
Quoted 4 lines
include/drm/drm_gpuva_mgr.h new file +4
-0+#define drm_gpuva_manager_set_ext_lock (mgr, lock) \ + (mgr)->ext_lock = &(lock)->dep_map +#else +typedef struct {/* nothing */ }lockdep_map_p ;
lockdep_map_pconflicts with an identical typedef inmaple_tree.hwhenCONFIG_LOCKDEPis not set (it's being pulled in bymm.hindrm_vma_manager.h). I'll just comment the line out for now.
Good catch! I got this trick from maple_tree.h and intended to move it to the lockdep header in a separate patch to avoid such collisions. Unfortunately, I forgot about it. Gonna fix it up.
- Danilo
Quoted 5 lines
Quoted 5 lines
include/drm/drm_gpuva_mgr.h new file +4
-0+#define drm_gpuva_manager_ext_assert_held (mgr) do { (void)(mgr); } while (0) +#define drm_gpuva_manager_set_ext_lock (mgr, lock) do { } while (0) +#endif +
Re: [PATCH drm-next v6 02/13] drm: manager to keep track of GPUs VA mappings
In reply to Danilo KrummrichQuoted 6 lines · On Fri, 30 Jun 2023 00:25:18 +0200 Danilo Krummrich <[email protected]> wrote:
include/drm/drm_gpuva_mgr.h new file +6
-0+#ifdef CONFIG_LOCKDEP +typedef struct lockdep_map *lockdep_map_p ; +#define drm_gpuva_manager_ext_assert_held (mgr) \ + lockdep_assert(lock_is_held((mgr)->ext_lock) != LOCK_STATE_NOT_HELD) +/** +* drm_gpuva_manager_set_ext_lock - set the external lock according to
include/drm/drm_gpuva_mgr.h line 203
Quoted 10 lines
include/drm/drm_gpuva_mgr.h new file +10
-0+* @DRM_GPUVA_MANAGER_LOCK_EXTERN +* @mgr: the &drm_gpuva_manager to set the lock for +* @lock: the lock to set +* +* If @DRM_GPUVA_MANAGER_LOCK_EXTERN is set, drivers need to call this function +* to provide the lock used to lock linking and unlinking of &drm_gpuvas to the +* &drm_gem_objects GPUVA list. +*/ +#define drm_gpuva_manager_set_ext_lock (mgr, lock) \ + (mgr)->ext_lock = &(lock)->dep_map
Okay, so, IIUC, this is the lock protecting the GEM's active mappings list, meaning the lock is likely to be attached to the GEM object. Are we expected to call drm_gpuva_manager_set_ext_lock() every time we call drm_gpuva_[un]link(), or are we supposed to have some lock at the device level serializing all drm_gpuva_[un]link() calls across VMs? The later doesn't sound like a good option to me, and the former feels a bit weird. I'm wondering if we shouldn't just drop this assert_held() check when DRM_GPUVA_MANAGER_LOCK_EXTERN is set. Alternatively, we could say that any driver wanting to use a custom lock (which is basically all drivers modifying the VA space asynchronously in the ::run_job() path) has to provide its own variant of drm_gpuva_[un]link() (maybe with its own VA list too), which doesn't sound like a good idea either.
Quoted 5 lines
include/drm/drm_gpuva_mgr.h new file +5
-0+#else +typedef struct {/* nothing */ }lockdep_map_p ; +#define drm_gpuva_manager_ext_assert_held (mgr) do { (void)(mgr); } while (0) +#define drm_gpuva_manager_set_ext_lock (mgr, lock) do { } while (0) +#endif
Re: [PATCH drm-next v6 02/13] drm: manager to keep track of GPUs VA mappings
Quoted 24 lines · On Thu, 6 Jul 2023 20:26:42 +0200 Boris Brezillon…
Quoted 6 lines · On Fri, 30 Jun 2023 00:25:18 +0200 Danilo Krummrich <[email protected]> wrote:
(quoted diff) +6
-0+#ifdef CONFIG_LOCKDEP +typedef struct lockdep_map *lockdep_map_p; +#define drm_gpuva_manager_ext_assert_held(mgr) \ + lockdep_assert(lock_is_held((mgr)->ext_lock) != LOCK_STATE_NOT_HELD) +/** + * drm_gpuva_manager_set_ext_lock - set the external lock according toReview
Quoted 10 lines
(quoted diff) +10
-0+ * @DRM_GPUVA_MANAGER_LOCK_EXTERN + * @mgr: the &drm_gpuva_manager to set the lock for + * @lock: the lock to set + * + * If @DRM_GPUVA_MANAGER_LOCK_EXTERN is set, drivers need to call this function + * to provide the lock used to lock linking and unlinking of &drm_gpuvas to the + * &drm_gem_objects GPUVA list. + */ +#define drm_gpuva_manager_set_ext_lock(mgr, lock) \ + (mgr)->ext_lock = &(lock)->dep_mapOkay, so, IIUC, this is the lock protecting the GEM's active mappings list, meaning the lock is likely to be attached to the GEM object. Are we expected to call
drm_gpuva_manager_set_ext_lock()every time we call drm_gpuva_[un]link(), or are we supposed to have some lock at the
Quoted 8 lines
device level serializing all drm_gpuva_[un]link() calls across VMs? The later doesn't sound like a good option to me, and the former feels a bit weird. I'm wondering if we shouldn't just drop this
assert_held()check whenDRM_GPUVA_MANAGER_LOCK_EXTERNis set. Alternatively, we could say that any driver wanting to use a custom lock (which is basically all drivers modifying the VA space asynchronously in the ::run_job() path) has to provide its own variant of drm_gpuva_[un]link() (maybe with its own VA list too), which doesn't sound like a good idea either.
Or we could just attach the dep_map to drm_gem_object::gpuva::lock, and let drivers overload the default lock in their GEM creation function if they want to use a custom lock (see the following diff).
---
drivers/gpu/drm/drm_gpuva_mgr.c +4 -17
@@ -675,8 +675,7 @@ drm_gpuva_manager_init(struct drm_gpuva_manager *mgr,const char *name, u64 start_offset, u64 range, u64 reserve_offset, u64 reserve_range, - const struct drm_gpuva_fn_ops *ops, - enum drm_gpuva_manager_flags flags) + const struct drm_gpuva_fn_ops *ops) { mgr->rb.tree = RB_ROOT_CACHED;INIT_LIST_HEAD (&mgr->rb.list);
@@ -686,7 +685,6 @@ drm_gpuva_manager_init(struct drm_gpuva_manager *mgr, mgr->mm_range = range; mgr->name = name ? name :"unknown" ; - mgr->flags = flags; mgr->ops = ops;memset (&mgr->kernel_alloc_node,0 ,sizeof (struct drm_gpuva ));
@@ -822,16 +820,12 @@ EXPORT_SYMBOL(drm_gpuva_remove);void drm_gpuva_link (struct drm_gpuva *va) { -struct drm_gpuva_manager *mgr = va->mgr;struct drm_gem_object *obj = va->gem.obj;if (unlikely (!obj))return ; -if (drm_gpuva_manager_external_lock (mgr)) -drm_gpuva_manager_ext_assert_held (mgr); -else -dma_resv_assert_held (obj->resv); +drm_gem_gpuva_assert_lock_held (obj);list_add_tail (&va->gem.entry, &obj->gpuva.list); }
@@ -850,16 +844,12 @@ EXPORT_SYMBOL(drm_gpuva_link);void drm_gpuva_unlink (struct drm_gpuva *va) { -struct drm_gpuva_manager *mgr = va->mgr;struct drm_gem_object *obj = va->gem.obj;if (unlikely (!obj))return ; -if (drm_gpuva_manager_external_lock (mgr)) -drm_gpuva_manager_ext_assert_held (mgr); -else -dma_resv_assert_held (obj->resv); +drm_gem_gpuva_assert_lock_held (obj);list_del_init (&va->gem.entry); }
@@ -1680,10 +1670,7 @@ drm_gpuva_gem_unmap_ops_create(struct drm_gpuva_manager *mgr,struct drm_gpuva *va;int ret; -if (drm_gpuva_manager_external_lock (mgr)) -drm_gpuva_manager_ext_assert_held (mgr); -else -dma_resv_assert_held (obj->resv); +drm_gem_gpuva_assert_lock_held (obj); ops =kzalloc (sizeof (*ops),GFP_KERNEL );if (!ops)
include/drm/drm_gem.h +26 -1
@@ -387,10 +387,14 @@ struct drm_gem_object {* Provides the list of GPU VAs attached to this GEM object. * * Drivers should lock list accesses with the GEMs &dma_resv lock -* (&drm_gem_object.resv). +* (&drm_gem_object.resv) or a custom lock if one is provided. */ struct {struct list_head list; + +#ifdef CONFIG_LOCKDEP +struct lockdep_map *lock_dep_map; +#endif } gpuva;/**
@@ -540,6 +544,26 @@ unsigned long drm_gem_lru_scan(struct drm_gem_lru *lru,int drm_gem_evict (struct drm_gem_object *obj); +#ifdef CONFIG_LOCKDEP +/* +* drm_gem_gpuva_set_lock() - Set the lock protecting accesses to the gpuva list. +* @obj: the &drm_gem_object +* @lock: the lock used to protect the gpuva list. The locking primitive +* must contain a dep_map field. +* +* Call this if you're not proctecting access to the gpuva list +* with the resv lock, otherwise, drm_gem_gpuva_init() takes case +* of initializing the lock_dep_map for you. +*/ +#define drm_gem_gpuva_set_lock (obj, lock) \ + obj->gpuva.lock_dep_map = &(lock)->dep_map +#define drm_gem_gpuva_assert_lock_held (obj) \ + lockdep_assert(lock_is_held(obj->gpuva.lock_dep_map)) +#else +#define drm_gem_gpuva_set_lock (obj, lock) do {} while(0) +#define drm_gem_gpuva_assert_lock_held (obj) do {} while(0) +#endif +/** * drm_gem_gpuva_init - initialize the gpuva list of a GEM object * @obj: the &drm_gem_object
@@ -552,6 +576,7 @@ int drm_gem_evict(struct drm_gem_object *obj);static inline void drm_gem_gpuva_init (struct drm_gem_object *obj) {INIT_LIST_HEAD (&obj->gpuva.list); +drm_gem_gpuva_set_lock (obj, &obj->resv->lock.base); }/**
include/drm/drm_gpuva_mgr.h +1 -60
87 lines not on this page · raw message
Re: [PATCH drm-next v6 02/13] drm: manager to keep track of GPUs VA mappings
Quoted 36 lines · Boris Brezillon
Quoted 24 lines · On Thu, 6 Jul 2023 20:26:42 +0200 Boris Brezillon…
Quoted 6 lines · On Fri, 30 Jun 2023 00:25:18 +0200 Danilo Krummrich <[email protected]> wrote:
(quoted diff) +6
-0+#ifdef CONFIG_LOCKDEP +typedef struct lockdep_map *lockdep_map_p; +#define drm_gpuva_manager_ext_assert_held(mgr) \ + lockdep_assert(lock_is_held((mgr)->ext_lock) != LOCK_STATE_NOT_HELD) +/** + * drm_gpuva_manager_set_ext_lock - set the external lock according toReview
Quoted 10 lines
(quoted diff) +10
-0+ * @DRM_GPUVA_MANAGER_LOCK_EXTERN + * @mgr: the &drm_gpuva_manager to set the lock for + * @lock: the lock to set + * + * If @DRM_GPUVA_MANAGER_LOCK_EXTERN is set, drivers need to call this function + * to provide the lock used to lock linking and unlinking of &drm_gpuvas to the + * &drm_gem_objects GPUVA list. + */ +#define drm_gpuva_manager_set_ext_lock(mgr, lock) \ + (mgr)->ext_lock = &(lock)->dep_mapOkay, so, IIUC, this is the lock protecting the GEM's active mappings list, meaning the lock is likely to be attached to the GEM object. Are we expected to call
drm_gpuva_manager_set_ext_lock()every time we call drm_gpuva_[un]link(), or are we supposed to have some lock at theQuoted 8 lines
device level serializing all drm_gpuva_[un]link() calls across VMs? The later doesn't sound like a good option to me, and the former feels a bit weird. I'm wondering if we shouldn't just drop this
assert_held()check whenDRM_GPUVA_MANAGER_LOCK_EXTERNis set. Alternatively, we could say that any driver wanting to use a custom lock (which is basically all drivers modifying the VA space asynchronously in the ::run_job() path) has to provide its own variant of drm_gpuva_[un]link() (maybe with its own VA list too), which doesn't sound like a good idea either.
Quoted 3 lines
Or we could just attach the
dep_maptodrm_gem_object::gpuva::lock, and let drivers overload the default lock in their GEM creation function if they want to use a custom lock (see the following diff).
Uh, I like that. Will pick it up, thanks!
Quoted 223 lines
---
drivers/gpu/drm/drm_gpuva_mgr.c +4
-17@@ -675,8 +675,7 @@ drm_gpuva_manager_init(struct drm_gpuva_manager *mgr,const char *name, u64 start_offset, u64 range, u64 reserve_offset, u64 reserve_range, - const struct drm_gpuva_fn_ops *ops, - enum drm_gpuva_manager_flags flags) + const struct drm_gpuva_fn_ops *ops) { mgr->rb.tree = RB_ROOT_CACHED;INIT_LIST_HEAD (&mgr->rb.list);@@ -686,7 +685,6 @@ drm_gpuva_manager_init(struct drm_gpuva_manager *mgr, mgr->mm_range = range; mgr->name = name ? name :"unknown" ; - mgr->flags = flags; mgr->ops = ops;memset (&mgr->kernel_alloc_node,0 ,sizeof (struct drm_gpuva ));@@ -822,16 +820,12 @@ EXPORT_SYMBOL(drm_gpuva_remove);void drm_gpuva_link (struct drm_gpuva *va) { -struct drm_gpuva_manager *mgr = va->mgr;struct drm_gem_object *obj = va->gem.obj;if (unlikely (!obj))return ; -if (drm_gpuva_manager_external_lock (mgr)) -drm_gpuva_manager_ext_assert_held (mgr); -else -dma_resv_assert_held (obj->resv); +drm_gem_gpuva_assert_lock_held (obj);list_add_tail (&va->gem.entry, &obj->gpuva.list); }@@ -850,16 +844,12 @@ EXPORT_SYMBOL(drm_gpuva_link);void drm_gpuva_unlink (struct drm_gpuva *va) { -struct drm_gpuva_manager *mgr = va->mgr;struct drm_gem_object *obj = va->gem.obj;if (unlikely (!obj))return ; -if (drm_gpuva_manager_external_lock (mgr)) -drm_gpuva_manager_ext_assert_held (mgr); -else -dma_resv_assert_held (obj->resv); +drm_gem_gpuva_assert_lock_held (obj);list_del_init (&va->gem.entry); }@@ -1680,10 +1670,7 @@ drm_gpuva_gem_unmap_ops_create(struct drm_gpuva_manager *mgr,struct drm_gpuva *va;int ret; -if (drm_gpuva_manager_external_lock (mgr)) -drm_gpuva_manager_ext_assert_held (mgr); -else -dma_resv_assert_held (obj->resv); +drm_gem_gpuva_assert_lock_held (obj); ops =kzalloc (sizeof (*ops),GFP_KERNEL );if (!ops)include/drm/drm_gem.h +26
-1@@ -387,10 +387,14 @@ struct drm_gem_object {* Provides the list of GPU VAs attached to this GEM object. * * Drivers should lock list accesses with the GEMs &dma_resv lock -* (&drm_gem_object.resv). +* (&drm_gem_object.resv) or a custom lock if one is provided. */ struct {struct list_head list; + +#ifdef CONFIG_LOCKDEP +struct lockdep_map *lock_dep_map; +#endif } gpuva;/** @@ -540,6 +544,26 @@ unsigned long drm_gem_lru_scan(struct drm_gem_lru *lru,int drm_gem_evict (struct drm_gem_object *obj); +#ifdef CONFIG_LOCKDEP +/* +* drm_gem_gpuva_set_lock() - Set the lock protecting accesses to the gpuva list. +* @obj: the &drm_gem_object +* @lock: the lock used to protect the gpuva list. The locking primitive +* must contain a dep_map field. +* +* Call this if you're not proctecting access to the gpuva list +* with the resv lock, otherwise, drm_gem_gpuva_init() takes case +* of initializing the lock_dep_map for you. +*/ +#define drm_gem_gpuva_set_lock (obj, lock) \ + obj->gpuva.lock_dep_map = &(lock)->dep_map +#define drm_gem_gpuva_assert_lock_held (obj) \ + lockdep_assert(lock_is_held(obj->gpuva.lock_dep_map)) +#else +#define drm_gem_gpuva_set_lock (obj, lock) do {} while(0) +#define drm_gem_gpuva_assert_lock_held (obj) do {} while(0) +#endif +/** * drm_gem_gpuva_init - initialize the gpuva list of a GEM object * @obj: the &drm_gem_object @@ -552,6 +576,7 @@ int drm_gem_evict(struct drm_gem_object *obj);static inline void drm_gem_gpuva_init (struct drm_gem_object *obj) {INIT_LIST_HEAD (&obj->gpuva.list); +drm_gem_gpuva_set_lock (obj, &obj->resv->lock.base); }/** include/drm/drm_gpuva_mgr.h +1
-60@@ -185,44 +185,6 @@ static inline bool drm_gpuva_invalidated(struct drm_gpuva *va)return va->flags &DRM_GPUVA_INVALIDATED ; } -#ifdef CONFIG_LOCKDEP -typedef struct lockdep_map *lockdep_map_p ; -#define drm_gpuva_manager_ext_assert_held (mgr) \ - lockdep_assert(lock_is_held((mgr)->ext_lock) != LOCK_STATE_NOT_HELD) -/** -* drm_gpuva_manager_set_ext_lock - set the external lock according to -* @DRM_GPUVA_MANAGER_LOCK_EXTERN -* @mgr: the &drm_gpuva_manager to set the lock for -* @lock: the lock to set -* -* If @DRM_GPUVA_MANAGER_LOCK_EXTERN is set, drivers need to call this function -* to provide the lock used to lock linking and unlinking of &drm_gpuvas to the -* &drm_gem_objects GPUVA list. -*/ -#define drm_gpuva_manager_set_ext_lock (mgr, lock) \ - (mgr)->ext_lock = &(lock)->dep_map -#else -typedef struct {/* nothing */ }lockdep_map_p ; -#define drm_gpuva_manager_ext_assert_held (mgr) do { (void)(mgr); } while (0) -#define drm_gpuva_manager_set_ext_lock (mgr, lock) do { } while (0) -#endif - -/** -* enum drm_gpuva_manager_flags - the feature flags for the &drm_gpuva_manager -*/ -enum drm_gpuva_manager_flags { -/** -* @DRM_GPUVA_MANAGER_LOCK_EXTERN: -* -* Indicates the driver has it's own external lock for linking and -* unlinking &drm_gpuvas to the &drm_gem_objects GPUVA list. -* -* When setting this flag it is rquired to set a lock via -* drm_gpuva_set_ext_lock(). -*/ -DRM_GPUVA_MANAGER_LOCK_EXTERN = (1 <<0 ), -}; -/** * struct drm_gpuva_manager - DRM GPU VA Manager * @@ -241,11 +203,6 @@ struct drm_gpuva_manager {*/ const char *name; -/** -* @flags: the feature flags of the &drm_gpuva_manager -*/ -enum drm_gpuva_manager_flags flags; -/** * @mm_start: start of the VA space */ @@ -283,31 +240,15 @@ struct drm_gpuva_manager {* @ops: &drm_gpuva_fn_ops providing the split/merge steps to drivers */ const struct drm_gpuva_fn_ops *ops; - -/** -* @ext_lock: &lockdep_map according to @DRM_GPUVA_MANAGER_LOCK_EXTERN -*/ -lockdep_map_p ext_lock; };void drm_gpuva_manager_init (struct drm_gpuva_manager *mgr,const char *name,u64 start_offset,u64 range,u64 reserve_offset,u64 reserve_range, -const struct drm_gpuva_fn_ops *ops, -enum drm_gpuva_manager_flags flags); +const struct drm_gpuva_fn_ops *ops);void drm_gpuva_manager_destroy (struct drm_gpuva_manager *mgr); -/** -* drm_gpuva_manager_external_lock - indicates whether the -* @DRM_GPUVA_MANAGER_LOCK_EXTERN flag is set -* @mgr: the &drm_gpuva_manager to check the flag for -*/ -static inline bool drm_gpuva_manager_external_lock (struct drm_gpuva_manager *mgr) -{ -return mgr->flags &DRM_GPUVA_MANAGER_LOCK_EXTERN ; -} -/** * drm_gpuva_for_each_va_range - iternator to walk over a range of &drm_gpuvas * @va__: &drm_gpuva structure to assign to in each iteration step
Re: [PATCH drm-next v6 02/13] drm: manager to keep track of GPUs VA mappings
In reply to Danilo KrummrichQuoted 8 lines · On Fri, 30 Jun 2023 00:25:18 +0200 Danilo Krummrich <[email protected]> wrote:
include/drm/drm_gpuva_mgr.h new file +8
-0+/** +* drm_gpuva_for_each_va_range - iternator to walk over a range of &drm_gpuvas +* @va__: &drm_gpuva structure to assign to in each iteration step +* @mgr__: &drm_gpuva_manager to walk over +* @start__: starting offset, the first gpuva will overlap this +* @end__: ending offset, the last gpuva will start before this (but may +* overlap) +*
include/drm/drm_gpuva_mgr.h line 328
Quoted 10 lines
include/drm/drm_gpuva_mgr.h new file +10
-0+* This iterator walks over all &drm_gpuvas in the &drm_gpuva_manager that lie +* between @start__ and @end__. It is implemented similarly to list_for_each(), +* but is using the &drm_gpuva_manager's internal interval tree to accelerate +* the search for the starting &drm_gpuva, and hence isn't safe against removal +* of elements. It assumes that @end__ is within (or is the upper limit of) the +* &drm_gpuva_manager. This iterator does not skip over the &drm_gpuva_manager's +* @kernel_alloc_node. +*/ +#define drm_gpuva_for_each_va_range (va__, mgr__, start__, end__) \ + for (va__ = drm_gpuva_find_first((mgr__), (start__), (end__)); \
drm_gpuva_find_first() takes the range size as its last argument, not the range end:
for (va__ = drm_gpuva_find_first((mgr__), (start__), (end__) - (start__)); \
include/drm/drm_gpuva_mgr.h line 331
Quoted 3 lines
include/drm/drm_gpuva_mgr.h new file +3
-0+ va__ && (va__->va.addr < (end__)) && \ + !list_entry_is_head (va__, &(mgr__)->rb.list, rb.entry); \ + va__ = list_next_entry(va__, rb.entry))
If you define:
static inline struct drm_gpuva *drm_gpuva_next(struct drm_gpuva *va)
{
if (va && !list_is_last(&va->rb.entry, &va->mgr->rb.list))
return list_next_entry(va, rb.entry);
return NULL;
}the for loop becomes a bit more readable:
for (va__ = drm_gpuva_find_first((mgr__), (start__), (end__) - (start__)); \ va__ && (va__->va.addr < (end__)); \ va__ = drm_gpuva_next(va__))
Quoted 15 lines
include/drm/drm_gpuva_mgr.h new file +15
-0+ +/** +* drm_gpuva_for_each_va_range_safe - iternator to safely walk over a range of +* &drm_gpuvas +* @va__: &drm_gpuva to assign to in each iteration step +* @next__: another &drm_gpuva to use as temporary storage +* @mgr__: &drm_gpuva_manager to walk over +* @start__: starting offset, the first gpuva will overlap this +* @end__: ending offset, the last gpuva will start before this (but may +* overlap) +* +* This iterator walks over all &drm_gpuvas in the &drm_gpuva_manager that lie +* between @start__ and @end__. It is implemented similarly to +* list_for_each_safe(), but is using the &drm_gpuva_manager's internal interval +* tree to accelerate the search for the starting &drm_gpuva, and hence is safe
include/drm/drm_gpuva_mgr.h line 356
Quoted 10 lines
include/drm/drm_gpuva_mgr.h new file +10
-0+* against removal of elements. It assumes that @end__ is within (or is the +* upper limit of) the &drm_gpuva_manager. This iterator does not skip over the +* &drm_gpuva_manager's @kernel_alloc_node. +*/ +#define drm_gpuva_for_each_va_range_safe (va__, next__, mgr__, start__, end__) \ + for (va__ = drm_gpuva_find_first((mgr__), (start__), (end__)), \ + next__ = va ? list_next_entry(va__, rb.entry) : NULL; \ + va__ && (va__->va.addr < (end__)) && \ + !list_entry_is_head(va__, &(mgr__)->rb.list, rb.entry); \ + va__ = next__, next__ = list_next_entry(va__, rb.entry))
And this is the safe version using the drm_gpuva_next() helper:
for (va__ = drm_gpuva_find_first((mgr__), (start__), (end__) - (start__)), \ next__ = drm_gpuva_next(va__); \ va__ && (va__->va.addr < (end__)); \ va__ = next__, next__ = drm_gpuva_next(va__))
Those changes fixed an invalid pointer access I had in the sm_unmap() path.
Re: [PATCH drm-next v6 02/13] drm: manager to keep track of GPUs VA mappings
Quoted 25 lines · Boris Brezillon
Quoted 8 lines · On Fri, 30 Jun 2023 00:25:18 +0200 Danilo Krummrich <[email protected]> wrote:
include/drm/drm_gpuva_mgr.h new file +8
-0+/** +* drm_gpuva_for_each_va_range - iternator to walk over a range of &drm_gpuvas +* @va__: &drm_gpuva structure to assign to in each iteration step +* @mgr__: &drm_gpuva_manager to walk over +* @start__: starting offset, the first gpuva will overlap this +* @end__: ending offset, the last gpuva will start before this (but may +* overlap) +* include/drm/drm_gpuva_mgr.h line 328
Quoted 10 lines
include/drm/drm_gpuva_mgr.h new file +10
-0+* This iterator walks over all &drm_gpuvas in the &drm_gpuva_manager that lie +* between @start__ and @end__. It is implemented similarly to list_for_each(), +* but is using the &drm_gpuva_manager's internal interval tree to accelerate +* the search for the starting &drm_gpuva, and hence isn't safe against removal +* of elements. It assumes that @end__ is within (or is the upper limit of) the +* &drm_gpuva_manager. This iterator does not skip over the &drm_gpuva_manager's +* @kernel_alloc_node. +*/ +#define drm_gpuva_for_each_va_range (va__, mgr__, start__, end__) \ + for (va__ = drm_gpuva_find_first((mgr__), (start__), (end__)); \
drm_gpuva_find_first()takes the range size as its last argument, not the range end:
Quoted 2 lines
for (va__ = drm_gpuva_find_first((mgr__), (start__), (end__) - (start__)); \
Good catch! Originally this was
drm_gpuva_it_iter_first(&(mgr)->rb.tree, (start__), (end__) - 1)
but then I changed it since I did not want to expose the interval tree functions directly.
Quoted 9 lines
include/drm/drm_gpuva_mgr.h line 331
Quoted 4 lines
include/drm/drm_gpuva_mgr.h new file +3
-0+ va__ && (va__->va.addr < (end__)) && \ + !list_entry_is_head (va__, &(mgr__)->rb.list, rb.entry); \ + va__ = list_next_entry(va__, rb.entry))If you define:
static inline struct
drm_gpuva*drm_gpuva_next(struct drm_gpuva *va)
Quoted 7 lines
{ if (va && !list_is_last(&va->rb.entry, &va->mgr->rb.list)) return list_next_entry(va, rb.entry); return NULL; } >the for loop becomes a bit more readable:
Yes, it would. However, I don't want it to be confused with drm_gpuva_find_next(). Maybe I should rename the latter to something like drm_gpuva_find_next_neighbor() then.
Quoted 33 lines
for (va__ = drm_gpuva_find_first((mgr__), (start__), (end__) - (start__)); \ va__ && (va__->va.addr < (end__)); \ va__ = drm_gpuva_next(va__))Quoted 15 lines
include/drm/drm_gpuva_mgr.h new file +15
-0+ +/** +* drm_gpuva_for_each_va_range_safe - iternator to safely walk over a range of +* &drm_gpuvas +* @va__: &drm_gpuva to assign to in each iteration step +* @next__: another &drm_gpuva to use as temporary storage +* @mgr__: &drm_gpuva_manager to walk over +* @start__: starting offset, the first gpuva will overlap this +* @end__: ending offset, the last gpuva will start before this (but may +* overlap) +* +* This iterator walks over all &drm_gpuvas in the &drm_gpuva_manager that lie +* between @start__ and @end__. It is implemented similarly to +* list_for_each_safe(), but is using the &drm_gpuva_manager's internal interval +* tree to accelerate the search for the starting &drm_gpuva, and hence is safe include/drm/drm_gpuva_mgr.h line 356
Quoted 10 lines
include/drm/drm_gpuva_mgr.h new file +10
-0+* against removal of elements. It assumes that @end__ is within (or is the +* upper limit of) the &drm_gpuva_manager. This iterator does not skip over the +* &drm_gpuva_manager's @kernel_alloc_node. +*/ +#define drm_gpuva_for_each_va_range_safe (va__, next__, mgr__, start__, end__) \ + for (va__ = drm_gpuva_find_first((mgr__), (start__), (end__)), \ + next__ = va ? list_next_entry(va__, rb.entry) : NULL; \ + va__ && (va__->va.addr < (end__)) && \ + !list_entry_is_head(va__, &(mgr__)->rb.list, rb.entry); \ + va__ = next__, next__ = list_next_entry(va__, rb.entry))And this is the safe version using the
drm_gpuva_next()helper:
Quoted 8 lines
for (va__ = drm_gpuva_find_first((mgr__), (start__), (end__) - (start__)), \ next__ = drm_gpuva_next(va__); \ va__ && (va__->va.addr < (end__)); \ va__ = next__, next__ = drm_gpuva_next(va__))Those changes fixed an invalid pointer access I had in the
sm_unmap()path.
Sorry you did run into this bug.
- Danilo
Re: [PATCH drm-next v6 02/13] drm: manager to keep track of GPUs VA mappings
Quoted 12 lines · On Fri, 7 Jul 2023 14:41:23 +0200 Danilo Krummrich <[email protected]> wrote:
Quoted 12 lines
include/drm/drm_gpuva_mgr.h line 331
Quoted 3 lines
include/drm/drm_gpuva_mgr.h new file +3
-0+ va__ && (va__->va.addr < (end__)) && \ + !list_entry_is_head (va__, &(mgr__)->rb.list, rb.entry); \ + va__ = list_next_entry(va__, rb.entry))If you define:
static inline struct
drm_gpuva*drm_gpuva_next(struct drm_gpuva *va){ if (va && !list_is_last(&va->rb.entry, &va->mgr->rb.list)) return list_next_entry(va, rb.entry);
Quoted 7 lines
Quoted 3 lines
return NULL; } >the for loop becomes a bit more readable:
Yes, it would. However, I don't want it to be confused with
drm_gpuva_find_next(). Maybe I should rename the latter to something likedrm_gpuva_find_next_neighbor()then.
If you want to keep drm_gpuva_find_next(), feel free to rename/prefix the drm_gpuva_next() function. I was just posting it as a reference.
Quoted 39 lines
Quoted 39 lines
for (va__ = drm_gpuva_find_first((mgr__), (start__), (end__) - (start__)); \ va__ && (va__->va.addr < (end__)); \ va__ = drm_gpuva_next(va__))Quoted 15 lines
include/drm/drm_gpuva_mgr.h new file +15
-0+ +/** +* drm_gpuva_for_each_va_range_safe - iternator to safely walk over a range of +* &drm_gpuvas +* @va__: &drm_gpuva to assign to in each iteration step +* @next__: another &drm_gpuva to use as temporary storage +* @mgr__: &drm_gpuva_manager to walk over +* @start__: starting offset, the first gpuva will overlap this +* @end__: ending offset, the last gpuva will start before this (but may +* overlap) +* +* This iterator walks over all &drm_gpuvas in the &drm_gpuva_manager that lie +* between @start__ and @end__. It is implemented similarly to +* list_for_each_safe(), but is using the &drm_gpuva_manager's internal interval +* tree to accelerate the search for the starting &drm_gpuva, and hence is safe include/drm/drm_gpuva_mgr.h line 356
Quoted 10 lines
include/drm/drm_gpuva_mgr.h new file +10
-0+* against removal of elements. It assumes that @end__ is within (or is the +* upper limit of) the &drm_gpuva_manager. This iterator does not skip over the +* &drm_gpuva_manager's @kernel_alloc_node. +*/ +#define drm_gpuva_for_each_va_range_safe (va__, next__, mgr__, start__, end__) \ + for (va__ = drm_gpuva_find_first((mgr__), (start__), (end__)), \ + next__ = va ? list_next_entry(va__, rb.entry) : NULL; \ + va__ && (va__->va.addr < (end__)) && \ + !list_entry_is_head(va__, &(mgr__)->rb.list, rb.entry); \ + va__ = next__, next__ = list_next_entry(va__, rb.entry))And this is the safe version using the
drm_gpuva_next()helper:for (va__ = drm_gpuva_find_first((mgr__), (start__), (end__) - (start__)), \ next__ = drm_gpuva_next(va__); \ va__ && (va__->va.addr < (end__)); \ va__ = next__, next__ = drm_gpuva_next(va__))
Quoted 5 lines
Quoted 3 lines
Those changes fixed an invalid pointer access I had in the
sm_unmap()path.Sorry you did run into this bug.
No worries, that's what testing/debugging/reviewing is for. And I'm glad someone decided to work on this gpuva stuff so I don't have to code it myself, so that's the least I can do.
Re: [PATCH drm-next v6 02/13] drm: manager to keep track of GPUs VA mappings
Quoted 69 lines · Boris Brezillon
Quoted 12 lines · On Fri, 7 Jul 2023 14:41:23 +0200 Danilo Krummrich <[email protected]> wrote:
Quoted 12 lines
include/drm/drm_gpuva_mgr.h line 331
Quoted 3 lines
include/drm/drm_gpuva_mgr.h new file +3
-0+ va__ && (va__->va.addr < (end__)) && \ + !list_entry_is_head (va__, &(mgr__)->rb.list, rb.entry); \ + va__ = list_next_entry(va__, rb.entry))If you define:
static inline struct
drm_gpuva*drm_gpuva_next(struct drm_gpuva *va){ if (va && !list_is_last(&va->rb.entry, &va->mgr->rb.list)) return list_next_entry(va, rb.entry);Quoted 7 lines
Quoted 3 lines
return NULL; } >the for loop becomes a bit more readable:
Yes, it would. However, I don't want it to be confused with
drm_gpuva_find_next(). Maybe I should rename the latter to something likedrm_gpuva_find_next_neighbor()then.If you want to keep
drm_gpuva_find_next(), feel free to rename/prefix thedrm_gpuva_next()function. I was just posting it as a reference.Quoted 43 lines
Quoted 42 lines
for (va__ = drm_gpuva_find_first((mgr__), (start__), (end__) - (start__)); \ va__ && (va__->va.addr < (end__)); \ va__ = drm_gpuva_next(va__))Quoted 15 lines
include/drm/drm_gpuva_mgr.h new file +15
-0+ +/** +* drm_gpuva_for_each_va_range_safe - iternator to safely walk over a range of +* &drm_gpuvas +* @va__: &drm_gpuva to assign to in each iteration step +* @next__: another &drm_gpuva to use as temporary storage +* @mgr__: &drm_gpuva_manager to walk over +* @start__: starting offset, the first gpuva will overlap this +* @end__: ending offset, the last gpuva will start before this (but may +* overlap) +* +* This iterator walks over all &drm_gpuvas in the &drm_gpuva_manager that lie +* between @start__ and @end__. It is implemented similarly to +* list_for_each_safe(), but is using the &drm_gpuva_manager's internal interval +* tree to accelerate the search for the starting &drm_gpuva, and hence is safe include/drm/drm_gpuva_mgr.h line 356
Quoted 10 lines
include/drm/drm_gpuva_mgr.h new file +10
-0+* against removal of elements. It assumes that @end__ is within (or is the +* upper limit of) the &drm_gpuva_manager. This iterator does not skip over the +* &drm_gpuva_manager's @kernel_alloc_node. +*/ +#define drm_gpuva_for_each_va_range_safe (va__, next__, mgr__, start__, end__) \ + for (va__ = drm_gpuva_find_first((mgr__), (start__), (end__)), \ + next__ = va ? list_next_entry(va__, rb.entry) : NULL; \ + va__ && (va__->va.addr < (end__)) && \ + !list_entry_is_head(va__, &(mgr__)->rb.list, rb.entry); \ + va__ = next__, next__ = list_next_entry(va__, rb.entry))And this is the safe version using the
drm_gpuva_next()helper:for (va__ = drm_gpuva_find_first((mgr__), (start__), (end__) - (start__)), \ next__ = drm_gpuva_next(va__); \ va__ && (va__->va.addr < (end__)); \ va__ = next__, next__ = drm_gpuva_next(va__))Those changes fixed an invalid pointer access I had in the
sm_unmap()path.
Quoted 5 lines
Quoted 1 line
Sorry you did run into this bug.
No worries, that's what testing/debugging/reviewing is for. And I'm glad someone decided to work on this gpuva stuff so I don't have to code it myself, so that's the least I can do.
With Boris's changes this version works in Xe.
With that:
PATCH v6 3/13 drm: debugfs: provide infrastructure to dump a DRM GPU VA space
In reply to Danilo KrummrichThis commit adds a function to dump a DRM GPU VA space and a macro for drivers to register the struct drm_info_list 'gpuvas' entry.
Most likely, most drivers might maintain one DRM GPU VA space per struct drm_file, but there might also be drivers not having a fixed relation between DRM GPU VA spaces and a DRM core infrastructure, hence we need the indirection via the driver iterating it's maintained DRM GPU VA spaces.
--- drivers/gpu/drm/drm_debugfs.c | 40 +++++++++++++++++++++++++++++++++++ include/drm/drm_debugfs.h | 25 ++++++++++++++++++++++ 2 files changed, 65 insertions(+)
drivers/gpu/drm/drm_debugfs.c +40 -0
@@ -39,6 +39,7 @@#include <drm/drm_file.h> #include <drm/drm_gem.h> #include <drm/drm_managed.h> +#include <drm/drm_gpuva_mgr.h> #include "drm_crtc_internal.h" #include "drm_internal.h"
@@ -175,6 +176,45 @@ static const struct file_operations drm_debugfs_fops = { .release = single_release, }; +/** +* drm_debugfs_gpuva_info - dump the given DRM GPU VA space +* @m: pointer to the &seq_file to write +* @mgr: the &drm_gpuva_manager representing the GPU VA space +* +* Dumps the GPU VA mappings of a given DRM GPU VA manager. +* +* For each DRM GPU VA space drivers should call this function from their +* &drm_info_list's show callback. +* +* Returns: 0 on success, -ENODEV if the &mgr is not initialized +*/ +int drm_debugfs_gpuva_info (struct seq_file *m, +struct drm_gpuva_manager *mgr) +{ +struct drm_gpuva *va, *kva = &mgr->kernel_alloc_node; + +if (!mgr->name) +return -ENODEV ; + +seq_printf (m,"DRM GPU VA space (%s) [0x%016llx;0x%016llx]\n" , + mgr->name, mgr->mm_start, mgr->mm_start + mgr->mm_range); +seq_printf (m,"Kernel reserved node [0x%016llx;0x%016llx]\n" , + kva->va.addr, kva->va.addr + kva->va.range); +seq_puts (m,"\n" ); +seq_puts (m," VAs | start | range | end | object | object offset\n" ); +seq_puts (m,"-------------------------------------------------------------------------------------------------------------\n" ); +drm_gpuva_for_each_va (va, mgr) { +if (unlikely (va == kva)) +continue ; + +seq_printf (m," | 0x%016llx | 0x%016llx | 0x%016llx | 0x%016llx | 0x%016llx\n" , + va->va.addr, va->va.range, va->va.addr + va->va.range, + (u64 )va->gem.obj, va->gem.offset); + } + +return 0 ; +} +EXPORT_SYMBOL (drm_debugfs_gpuva_info);/** * drm_debugfs_create_files - Initialize a given set of debugfs files for DRM
include/drm/drm_debugfs.h +25 -0
@@ -34,6 +34,22 @@#include <linux/types.h> #include <linux/seq_file.h> + +#include <drm/drm_gpuva_mgr.h> + +/** +* DRM_DEBUGFS_GPUVA_INFO - &drm_info_list entry to dump a GPU VA space +* @show: the &drm_info_list's show callback +* @data: driver private data +* +* Drivers should use this macro to define a &drm_info_list entry to provide a +* debugfs file for dumping the GPU VA space regions and mappings. +* +* For each DRM GPU VA space drivers should call drm_debugfs_gpuva_info() from +* their @show callback. +*/ +#define DRM_DEBUGFS_GPUVA_INFO (show, data) {"gpuvas", show, DRIVER_GEM_GPUVA, data} +/** * struct drm_info_list - debugfs info list entry *
@@ -134,6 +150,9 @@ void drm_debugfs_add_file(struct drm_device *dev, const char *name,void drm_debugfs_add_files (struct drm_device *dev,const struct drm_debugfs_info *files,int count); + +int drm_debugfs_gpuva_info (struct seq_file *m, +struct drm_gpuva_manager *mgr);#else static inline void drm_debugfs_create_files(const struct drm_info_list *files, int count, struct dentry *root,
@@ -155,6 +174,12 @@ static inline void drm_debugfs_add_files(struct drm_device *dev, const struct drm_debugfs_info *files, int count) {} + +static inline int drm_debugfs_gpuva_info (struct seq_file *m, +struct drm_gpuva_manager *mgr) +{ +return 0 ; +}#endif #endif /* _DRM_DEBUGFS_H_ */
Signature · 1 line
2.41.0
PATCH v6 4/13 drm/nouveau: new VM_BIND uapi interfaces
In reply to Danilo KrummrichThis commit provides the interfaces for the new UAPI motivated by the Vulkan API. It allows user mode drivers (UMDs) to:
1) Initialize a GPU virtual address (VA) space via the new DRM_IOCTL_NOUVEAU_VM_INIT ioctl. UMDs can provide a kernel reserved VA area.
2) Bind and unbind GPU VA space mappings via the new DRM_IOCTL_NOUVEAU_VM_BIND ioctl.
3) Execute push buffers with the new DRM_IOCTL_NOUVEAU_EXEC ioctl.
Both, DRM_IOCTL_NOUVEAU_VM_BIND and DRM_IOCTL_NOUVEAU_EXEC support asynchronous processing with DRM syncobjs as synchronization mechanism.
The default DRM_IOCTL_NOUVEAU_VM_BIND is synchronous processing, DRM_IOCTL_NOUVEAU_EXEC supports asynchronous processing only.
--- Documentation/gpu/driver-uapi.rst | 8 ++ include/uapi/drm/nouveau_drm.h | 209 ++++++++++++++++++++++++++++++ 2 files changed, 217 insertions(+)
Documentation/gpu/driver-uapi.rst +8 -0
@@ -6,3 +6,11 @@ drm/i915 uAPI ============= .. kernel-doc:: include/uapi/drm/i915_drm.h + +drm/nouveau uAPI +================ + +VM_BIND / EXEC uAPI +------------------- + +.. kernel-doc:: include/uapi/drm/nouveau_drm.h
include/uapi/drm/nouveau_drm.h +209 -0
227 lines not on this page · raw message
Signature · 1 line
2.41.0
PATCH v6 5/13 drm/nouveau: get vmm via nouveau_cli_vmm()
In reply to Danilo KrummrichProvide a getter function for the client's current vmm context. Since we'll add a new (u)vmm context for UMD bindings in subsequent commits, this will keep the code clean.
--- drivers/gpu/drm/nouveau/nouveau_bo.c | 2 +- drivers/gpu/drm/nouveau/nouveau_chan.c | 2 +- drivers/gpu/drm/nouveau/nouveau_drv.h | 9 +++++++++ drivers/gpu/drm/nouveau/nouveau_gem.c | 6 +++--- 4 files changed, 14 insertions(+), 5 deletions(-)
drivers/gpu/drm/nouveau/nouveau_bo.c +1 -1
@@ -204,7 +204,7 @@ nouveau_bo_alloc(struct nouveau_cli *cli, u64 *size, int *align, u32 domain,struct nouveau_drm *drm = cli->drm;struct nouveau_bo *nvbo;struct nvif_mmu *mmu = &cli->mmu; -struct nvif_vmm *vmm = cli->svm.cli ? &cli->svm.vmm : &cli->vmm.vmm; +struct nvif_vmm *vmm = &nouveau_cli_vmm (cli)->vmm;int i, pi =-1 ;if (!*size) {
drivers/gpu/drm/nouveau/nouveau_chan.c +1 -1
@@ -148,7 +148,7 @@ nouveau_channel_prep(struct nouveau_drm *drm, struct nvif_device *device, chan->device = device; chan->drm = drm; - chan->vmm = cli->svm.cli ? &cli->svm : &cli->vmm; + chan->vmm =nouveau_cli_vmm (cli);atomic_set (&chan->killed,0 );/* allocate memory for dma push buffer */
drivers/gpu/drm/nouveau/nouveau_drv.h +9 -0
@@ -112,6 +112,15 @@ struct nouveau_cli_work {struct dma_fence_cb cb; }; +static inline struct nouveau_vmm * +nouveau_cli_vmm (struct nouveau_cli *cli) +{ +if (cli->svm.cli) +return &cli->svm; + +return &cli->vmm; +} +void nouveau_cli_work_queue (struct nouveau_cli *,struct dma_fence *,struct nouveau_cli_work *);
drivers/gpu/drm/nouveau/nouveau_gem.c +3 -3
@@ -103,7 +103,7 @@ nouveau_gem_object_open(struct drm_gem_object *gem, struct drm_file *file_priv)struct nouveau_bo *nvbo =nouveau_gem_object (gem);struct nouveau_drm *drm =nouveau_bdev (nvbo->bo.bdev);struct device *dev = drm->dev->dev; -struct nouveau_vmm *vmm = cli->svm.cli ? &cli->svm : &cli->vmm; +struct nouveau_vmm *vmm =nouveau_cli_vmm (cli);struct nouveau_vma *vma;int ret;
@@ -180,7 +180,7 @@ nouveau_gem_object_close(struct drm_gem_object *gem, struct drm_file *file_priv)struct nouveau_bo *nvbo =nouveau_gem_object (gem);struct nouveau_drm *drm =nouveau_bdev (nvbo->bo.bdev);struct device *dev = drm->dev->dev; -struct nouveau_vmm *vmm = cli->svm.cli ? &cli->svm : & cli->vmm; +struct nouveau_vmm *vmm =nouveau_cli_vmm (cli);struct nouveau_vma *vma;int ret;
@@ -269,7 +269,7 @@ nouveau_gem_info(struct drm_file *file_priv, struct drm_gem_object *gem, {struct nouveau_cli *cli =nouveau_cli (file_priv);struct nouveau_bo *nvbo =nouveau_gem_object (gem); -struct nouveau_vmm *vmm = cli->svm.cli ? &cli->svm : &cli->vmm; +struct nouveau_vmm *vmm =nouveau_cli_vmm (cli);struct nouveau_vma *vma;if (is_power_of_2 (nvbo->valid_domains))
Signature · 1 line
2.41.0
PATCH v6 6/13 drm/nouveau: bo: initialize GEM GPU VA interface
In reply to Danilo KrummrichInitialize the GEM's DRM GPU VA manager interface in preparation for the (u)vmm implementation, provided by subsequent commits, to make use of it.
--- drivers/gpu/drm/nouveau/nouveau_bo.c | 3 +++ 1 file changed, 3 insertions(+)
drivers/gpu/drm/nouveau/nouveau_bo.c +3 -0
@@ -215,11 +215,14 @@ nouveau_bo_alloc(struct nouveau_cli *cli, u64 *size, int *align, u32 domain, nvbo =kzalloc (sizeof (struct nouveau_bo ),GFP_KERNEL );if (!nvbo)return ERR_PTR (-ENOMEM ); +INIT_LIST_HEAD (&nvbo->head);INIT_LIST_HEAD (&nvbo->entry);INIT_LIST_HEAD (&nvbo->vma_list); nvbo->bo.bdev = &drm->ttm.bdev; +drm_gem_gpuva_init (&nvbo->bo.base); +/* This is confusing, and doesn't actually mean we want an uncached * mapping, but is what NOUVEAU_GEM_DOMAIN_COHERENT gets translated * into in nouveau_gem_new().
Signature · 1 line
2.41.0
PATCH v6 7/13 drm/nouveau: move usercopy helpers to nouveau_drv.h
In reply to Danilo KrummrichMove the usercopy helpers to a common driver header file to make it usable for the new API added in subsequent commits.
--- drivers/gpu/drm/nouveau/nouveau_drv.h | 26 ++++++++++++++++++++++++++ drivers/gpu/drm/nouveau/nouveau_gem.c | 26 -------------------------- 2 files changed, 26 insertions(+), 26 deletions(-)
drivers/gpu/drm/nouveau/nouveau_drv.h +26 -0
@@ -130,6 +130,32 @@ nouveau_cli(struct drm_file *fpriv)return fpriv ? fpriv->driver_priv :NULL ; } +static inline void +u_free (void *addr) +{ +kvfree (addr); +} + +static inline void * +u_memcpya (uint64_t user,unsigned nmemb,unsigned size) +{ +void *mem; +void __user *userptr = (void __force __user *)(uintptr_t )user; + + size *= nmemb; + + mem =kvmalloc (size,GFP_KERNEL ); +if (!mem) +return ERR_PTR (-ENOMEM ); + +if (copy_from_user (mem, userptr, size)) { +u_free (mem); +return ERR_PTR (-EFAULT ); + } + +return mem; +} +#include <nvif/object.h> #include <nvif/parent.h>
drivers/gpu/drm/nouveau/nouveau_gem.c +0 -26
@@ -613,32 +613,6 @@ nouveau_gem_pushbuf_validate(struct nouveau_channel *chan,return 0 ; } -static inline void -u_free (void *addr) -{ -kvfree (addr); -} - -static inline void * -u_memcpya (uint64_t user,unsigned nmemb,unsigned size) -{ -void *mem; -void __user *userptr = (void __force __user *)(uintptr_t )user; - - size *= nmemb; - - mem =kvmalloc (size,GFP_KERNEL ); -if (!mem) -return ERR_PTR (-ENOMEM ); - -if (copy_from_user (mem, userptr, size)) { -u_free (mem); -return ERR_PTR (-EFAULT ); - } - -return mem; -} - static int nouveau_gem_pushbuf_reloc_apply(struct nouveau_cli *cli, struct drm_nouveau_gem_pushbuf *req,
Signature · 1 line
2.41.0
PATCH v6 8/13 drm/nouveau: fence: separate fence alloc and emit
In reply to Danilo KrummrichThe new (VM_BIND) UAPI exports DMA fences through DRM syncobjs. Hence, in order to emit fences within DMA fence signalling critical sections (e.g. as typically done in the DRM GPU schedulers run_job() callback) we need to separate fence allocation and fence emitting.
--- drivers/gpu/drm/nouveau/dispnv04/crtc.c | 9 ++++- drivers/gpu/drm/nouveau/nouveau_bo.c | 52 +++++++++++++++---------- drivers/gpu/drm/nouveau/nouveau_chan.c | 6 ++- drivers/gpu/drm/nouveau/nouveau_dmem.c | 9 +++-- drivers/gpu/drm/nouveau/nouveau_fence.c | 16 +++----- drivers/gpu/drm/nouveau/nouveau_fence.h | 3 +- drivers/gpu/drm/nouveau/nouveau_gem.c | 5 ++- 7 files changed, 59 insertions(+), 41 deletions(-)
drivers/gpu/drm/nouveau/dispnv04/crtc.c +8 -1
@@ -1122,11 +1122,18 @@ nv04_page_flip_emit(struct nouveau_channel *chan,PUSH_NVSQ (push,NV_SW ,NV_SW_PAGE_FLIP ,0x00000000 );PUSH_KICK (push); - ret =nouveau_fence_new (chan, false, pfence); + ret =nouveau_fence_new (pfence);if (ret) goto fail; + ret =nouveau_fence_emit (*pfence, chan); +if (ret) + goto fail_fence_unref; +return 0 ; + +fail_fence_unref: +nouveau_fence_unref (pfence); fail:spin_lock_irqsave (&dev->event_lock, flags);list_del (&s->head);
drivers/gpu/drm/nouveau/nouveau_bo.c +31 -21
60 lines not on this page · raw message
drivers/gpu/drm/nouveau/nouveau_chan.c +4 -2
@@ -62,9 +62,11 @@ nouveau_channel_idle(struct nouveau_channel *chan)struct nouveau_fence *fence =NULL ;int ret; - ret =nouveau_fence_new (chan, false, &fence); + ret =nouveau_fence_new (&fence);if (!ret) { - ret =nouveau_fence_wait (fence, false, false); + ret =nouveau_fence_emit (fence, chan); +if (!ret) + ret =nouveau_fence_wait (fence, false, false);nouveau_fence_unref (&fence); }
drivers/gpu/drm/nouveau/nouveau_dmem.c +6 -3
@@ -209,7 +209,8 @@ static vm_fault_t nouveau_dmem_migrate_to_ram(struct vm_fault *vmf) goto done; } -nouveau_fence_new (dmem->migrate.chan, false, &fence); +if (!nouveau_fence_new (&fence)) +nouveau_fence_emit (fence, dmem->migrate.chan);migrate_vma_pages (&args);nouveau_dmem_fence_done (&fence);dma_unmap_page (drm->dev->dev, dma_addr,PAGE_SIZE ,DMA_BIDIRECTIONAL );
@@ -402,7 +403,8 @@ nouveau_dmem_evict_chunk(struct nouveau_dmem_chunk *chunk) } } -nouveau_fence_new (chunk->drm->dmem->migrate.chan, false, &fence); +if (!nouveau_fence_new (&fence)) +nouveau_fence_emit (fence, chunk->drm->dmem->migrate.chan);migrate_device_pages (src_pfns, dst_pfns, npages);nouveau_dmem_fence_done (&fence);migrate_device_finalize (src_pfns, dst_pfns, npages);
@@ -675,7 +677,8 @@ static void nouveau_dmem_migrate_chunk(struct nouveau_drm *drm, addr +=PAGE_SIZE ; } -nouveau_fence_new (drm->dmem->migrate.chan, false, &fence); +if (!nouveau_fence_new (&fence)) +nouveau_fence_emit (fence, chunk->drm->dmem->migrate.chan);migrate_vma_pages (args);nouveau_dmem_fence_done (&fence);nouveau_pfns_map (svmm, args->vma->vm_mm, args->start, pfns, i);
drivers/gpu/drm/nouveau/nouveau_fence.c +5 -11
@@ -210,6 +210,9 @@ nouveau_fence_emit(struct nouveau_fence *fence, struct nouveau_channel *chan)struct nouveau_fence_priv *priv = (void *)chan->drm->fence;int ret; +if (unlikely (!chan->fence)) +return -ENODEV ; + fence->channel = chan; fence->timeout = jiffies + (15 *HZ );
@@ -396,25 +399,16 @@ nouveau_fence_unref(struct nouveau_fence **pfence) }int -nouveau_fence_new (struct nouveau_channel *chan,bool sysmem, -struct nouveau_fence **pfence) +nouveau_fence_new (struct nouveau_fence **pfence) {struct nouveau_fence *fence; -int ret =0 ; - -if (unlikely (!chan->fence)) -return -ENODEV ; fence =kzalloc (sizeof (*fence),GFP_KERNEL );if (!fence)return -ENOMEM ; - ret =nouveau_fence_emit (fence, chan); -if (ret) -nouveau_fence_unref (&fence); - *pfence = fence; -return ret; +return 0 ; }static const char *nouveau_fence_get_get_driver_name (struct dma_fence *fence)
drivers/gpu/drm/nouveau/nouveau_fence.h +1 -2
@@ -17,8 +17,7 @@ struct nouveau_fence {unsigned long timeout; }; -int nouveau_fence_new (struct nouveau_channel *,bool sysmem, -struct nouveau_fence **); +int nouveau_fence_new (struct nouveau_fence **);void nouveau_fence_unref (struct nouveau_fence **);int nouveau_fence_emit (struct nouveau_fence *,struct nouveau_channel *);
drivers/gpu/drm/nouveau/nouveau_gem.c +4 -1
@@ -873,8 +873,11 @@ nouveau_gem_ioctl_pushbuf(struct drm_device *dev, void *data, } } - ret =nouveau_fence_new (chan, false, &fence); + ret =nouveau_fence_new (&fence); +if (!ret) + ret =nouveau_fence_emit (fence, chan);if (ret) { +nouveau_fence_unref (&fence);NV_PRINTK (err, cli,"error fencing pushbuf: %d\n" , ret);WIND_RING (chan); goto out;
Signature · 1 line
2.41.0
PATCH v6 9/13 drm/nouveau: fence: fail to emit when fence context is killed
In reply to Danilo KrummrichThe new VM_BIND UAPI implementation introduced in subsequent commits will allow asynchronous jobs processing push buffers and emitting fences.
If a fence context is killed, e.g. due to a channel fault, jobs which are already queued for execution might still emit new fences. In such a case a job would hang forever.
To fix that, fail to emit a new fence on a killed fence context with -ENODEV to unblock the job.
--- drivers/gpu/drm/nouveau/nouveau_fence.c | 7 +++++++ drivers/gpu/drm/nouveau/nouveau_fence.h | 2 +- 2 files changed, 8 insertions(+), 1 deletion(-)
drivers/gpu/drm/nouveau/nouveau_fence.c +7 -0
@@ -96,6 +96,7 @@ nouveau_fence_context_kill(struct nouveau_fence_chan *fctx, int error)if (nouveau_fence_signal (fence))nvif_event_block (&fctx->event); } + fctx->killed =1 ;spin_unlock_irqrestore (&fctx->lock, flags); }
@@ -229,6 +230,12 @@ nouveau_fence_emit(struct nouveau_fence *fence, struct nouveau_channel *chan)dma_fence_get (&fence->base);spin_lock_irq (&fctx->lock); +if (unlikely (fctx->killed)) { +spin_unlock_irq (&fctx->lock); +dma_fence_put (&fence->base); +return -ENODEV ; + } +if (nouveau_fence_update (chan, fctx))nvif_event_block (&fctx->event);
drivers/gpu/drm/nouveau/nouveau_fence.h +1 -1
@@ -44,7 +44,7 @@ struct nouveau_fence_chan {char name[32 ];struct nvif_event event; -int notify_ref, dead; +int notify_ref, dead, killed; };struct nouveau_fence_priv {
Signature · 1 line
2.41.0
PATCH v6 10/13 drm/nouveau: chan: provide nouveau_channel_kill()
In reply to Danilo KrummrichThe new VM_BIND UAPI implementation introduced in subsequent commits will allow asynchronous jobs processing push buffers and emitting fences.
If a job times out, we need a way to recover from this situation. For now, simply kill the channel to unblock all hung up jobs and signal userspace that the device is dead on the next EXEC or VM_BIND ioctl.
--- drivers/gpu/drm/nouveau/nouveau_chan.c | 14 +++++++++++--- drivers/gpu/drm/nouveau/nouveau_chan.h | 1 + 2 files changed, 12 insertions(+), 3 deletions(-)
drivers/gpu/drm/nouveau/nouveau_chan.c +11 -3
@@ -40,6 +40,14 @@ MODULE_PARM_DESC(vram_pushbuf, "Create DMA push buffers in VRAM");int nouveau_vram_pushbuf;module_param_named (vram_pushbuf, nouveau_vram_pushbuf, int,0400 ); +void +nouveau_channel_kill (struct nouveau_channel *chan) +{ +atomic_set (&chan->killed,1 ); +if (chan->fence) +nouveau_fence_context_kill (chan->fence, -ENODEV ); +} +static int nouveau_channel_killed (struct nvif_event *event,void *repv,u32 repc) {
@@ -47,9 +55,9 @@ nouveau_channel_killed(struct nvif_event *event, void *repv, u32 repc)struct nouveau_cli *cli = (void *)chan->user.client;NV_PRINTK (warn, cli,"channel %d killed!\n" , chan->chid); -atomic_set (&chan->killed,1 ); -if (chan->fence) -nouveau_fence_context_kill (chan->fence, -ENODEV ); + +if (unlikely (!atomic_read (&chan->killed))) +nouveau_channel_kill (chan);return NVIF_EVENT_DROP ; }
drivers/gpu/drm/nouveau/nouveau_chan.h +1 -0
@@ -65,6 +65,7 @@ int nouveau_channel_new(struct nouveau_drm *, struct nvif_device *, bool priv, u32 vram, u32 gart, struct nouveau_channel **); void nouveau_channel_del(struct nouveau_channel **);int nouveau_channel_idle (struct nouveau_channel *); +void nouveau_channel_kill (struct nouveau_channel *);extern int nouveau_vram_pushbuf;
Signature · 1 line
2.41.0
PATCH v6 11/13 drm/nouveau: nvkm/vmm: implement raw ops to manage uvmm
In reply to Danilo KrummrichThe new VM_BIND UAPI uses the DRM GPU VA manager to manage the VA space. Hence, we a need a way to manipulate the MMUs page tables without going through the internal range allocator implemented by nvkm/vmm.
This patch adds a raw interface for nvkm/vmm to pass the resposibility for managing the address space and the corresponding map/unmap/sparse operations to the upper layers.
--- drivers/gpu/drm/nouveau/include/nvif/if000c.h | 26 ++- drivers/gpu/drm/nouveau/include/nvif/vmm.h | 19 +- .../gpu/drm/nouveau/include/nvkm/subdev/mmu.h | 20 +- drivers/gpu/drm/nouveau/nouveau_svm.c | 2 +- drivers/gpu/drm/nouveau/nouveau_vmm.c | 4 +- drivers/gpu/drm/nouveau/nvif/vmm.c | 100 +++++++- .../gpu/drm/nouveau/nvkm/subdev/mmu/uvmm.c | 213 ++++++++++++++++-- drivers/gpu/drm/nouveau/nvkm/subdev/mmu/vmm.c | 197 ++++++++++++---- drivers/gpu/drm/nouveau/nvkm/subdev/mmu/vmm.h | 25 ++ .../drm/nouveau/nvkm/subdev/mmu/vmmgf100.c | 16 +- .../drm/nouveau/nvkm/subdev/mmu/vmmgp100.c | 16 +- .../gpu/drm/nouveau/nvkm/subdev/mmu/vmmnv50.c | 27 ++- 12 files changed, 566 insertions(+), 99 deletions(-)
drivers/gpu/drm/nouveau/include/nvif/if000c.h +25 -1
@@ -3,7 +3,10 @@struct nvif_vmm_v0 {__u8 version;__u8 page_nr; -__u8 managed; +#define NVIF_VMM_V0_TYPE_UNMANAGED 0x00 +#define NVIF_VMM_V0_TYPE_MANAGED 0x01 +#define NVIF_VMM_V0_TYPE_RAW 0x02 +__u8 type;__u8 pad03[5 ];__u64 addr;__u64 size;
@@ -17,6 +20,7 @@ struct nvif_vmm_v0 {#define NVIF_VMM_V0_UNMAP 0x04#define NVIF_VMM_V0_PFNMAP 0x05#define NVIF_VMM_V0_PFNCLR 0x06 +#define NVIF_VMM_V0_RAW 0x07#define NVIF_VMM_V0_MTHD (i) ((i) + 0x80)struct nvif_vmm_page_v0 {
@@ -66,6 +70,26 @@ struct nvif_vmm_unmap_v0 {__u64 addr; }; +struct nvif_vmm_raw_v0 { +__u8 version; +#define NVIF_VMM_RAW_V0_GET 0x0 +#define NVIF_VMM_RAW_V0_PUT 0x1 +#define NVIF_VMM_RAW_V0_MAP 0x2 +#define NVIF_VMM_RAW_V0_UNMAP 0x3 +#define NVIF_VMM_RAW_V0_SPARSE 0x4 +__u8 op; +__u8 sparse; +__u8 ref; +__u8 shift; +__u32 argc; +__u8 pad01[7 ]; +__u64 addr; +__u64 size; +__u64 offset; +__u64 memory; +__u64 argv; +}; +struct nvif_vmm_pfnmap_v0 {__u8 version;__u8 page;
drivers/gpu/drm/nouveau/include/nvif/vmm.h +17 -2
@@ -4,6 +4,12 @@struct nvif_mem ;struct nvif_mmu ; +enum nvif_vmm_type { +UNMANAGED , +MANAGED , +RAW , +}; +enum nvif_vmm_get {ADDR ,PTES ,
@@ -30,8 +36,9 @@ struct nvif_vmm {int page_nr; }; -int nvif_vmm_ctor (struct nvif_mmu *,const char *name,s32 oclass,bool managed, -u64 addr,u64 size,void *argv,u32 argc,struct nvif_vmm *); +int nvif_vmm_ctor (struct nvif_mmu *,const char *name,s32 oclass, +enum nvif_vmm_type ,u64 addr,u64 size,void *argv,u32 argc, +struct nvif_vmm *);void nvif_vmm_dtor (struct nvif_vmm *);int nvif_vmm_get (struct nvif_vmm *,enum nvif_vmm_get ,bool sparse,u8 page,u8 align,u64 size,struct nvif_vma *);
@@ -39,4 +46,12 @@ void nvif_vmm_put(struct nvif_vmm *, struct nvif_vma *);int nvif_vmm_map (struct nvif_vmm *,u64 addr,u64 size,void *argv,u32 argc,struct nvif_mem *,u64 offset);int nvif_vmm_unmap (struct nvif_vmm *,u64 ); + +int nvif_vmm_raw_get (struct nvif_vmm *vmm,u64 addr,u64 size,u8 shift); +int nvif_vmm_raw_put (struct nvif_vmm *vmm,u64 addr,u64 size,u8 shift); +int nvif_vmm_raw_map (struct nvif_vmm *vmm,u64 addr,u64 size,u8 shift, +void *argv,u32 argc,struct nvif_mem *mem,u64 offset); +int nvif_vmm_raw_unmap (struct nvif_vmm *vmm,u64 addr,u64 size, +u8 shift,bool sparse); +int nvif_vmm_raw_sparse (struct nvif_vmm *vmm,u64 addr,u64 size,bool ref);#endif
drivers/gpu/drm/nouveau/include/nvkm/subdev/mmu.h +19 -1
@@ -17,6 +17,7 @@ struct nvkm_vma { bool part:1; /* Region was split from an allocated region by map(). */ bool busy:1; /* Region busy (for temporarily preventing user access). */ bool mapped:1; /* Region contains valid pages. */ + bool no_comp:1; /* Force no memory compression. */struct nvkm_memory *memory;/* Memory currently mapped into VMA. */ struct nvkm_tags *tags;/* Compression tag reference. */ };
@@ -27,10 +28,26 @@ struct nvkm_vmm {const char *name;u32 debug;struct kref kref; -struct mutex mutex; + +struct { +struct mutex vmm; +struct mutex ref; +struct mutex map; + } mutex;u64 start;u64 limit; +struct { +struct { +u64 addr; +u64 size; + } p; +struct { +u64 addr; +u64 size; + } n; +bool raw; + } managed;struct nvkm_vmm_pt *pd;struct list_head join;
@@ -70,6 +87,7 @@ struct nvkm_vmm_map {const struct nvkm_vmm_page *page; +bool no_comp;struct nvkm_tags *tags;u64 next;u64 type;
drivers/gpu/drm/nouveau/nouveau_svm.c +1 -1
@@ -350,7 +350,7 @@ nouveau_svmm_init(struct drm_device *dev, void *data,* VMM instead of the standard one. */ ret = nvif_vmm_ctor(&cli->mmu, "svmVmm", - cli->vmm.vmm.object.oclass, true, + cli->vmm.vmm.object.oclass, MANAGED, args->unmanaged_addr, args->unmanaged_size, &(struct gp100_vmm_v0) { .fault_replay = true,
drivers/gpu/drm/nouveau/nouveau_vmm.c +2 -2
@@ -128,8 +128,8 @@ nouveau_vmm_fini(struct nouveau_vmm *vmm)int nouveau_vmm_init (struct nouveau_cli *cli,s32 oclass,struct nouveau_vmm *vmm) { -int ret =nvif_vmm_ctor (&cli->mmu,"drmVmm" , oclass, false,PAGE_SIZE , -0 ,NULL ,0 , &vmm->vmm); +int ret =nvif_vmm_ctor (&cli->mmu,"drmVmm" , oclass,UNMANAGED , +PAGE_SIZE ,0 ,NULL ,0 , &vmm->vmm);if (ret)return ret;
drivers/gpu/drm/nouveau/nvif/vmm.c +97 -3
120 lines not on this page · raw message
drivers/gpu/drm/nouveau/nvkm/subdev/mmu/uvmm.c +195 -18
324 lines not on this page · raw message
drivers/gpu/drm/nouveau/nvkm/subdev/mmu/vmm.c +152 -45
306 lines not on this page · raw message
drivers/gpu/drm/nouveau/nvkm/subdev/mmu/vmm.h +25 -0
@@ -163,6 +163,7 @@ int nvkm_vmm_new_(const struct nvkm_vmm_func *, struct nvkm_mmu *, u32 pd_header, bool managed, u64 addr, u64 size, struct lock_class_key *, const char *name, struct nvkm_vmm **); +struct nvkm_vma *nvkm_vma_new (u64 addr,u64 size);struct nvkm_vma *nvkm_vmm_node_search (struct nvkm_vmm *,u64 addr);struct nvkm_vma *nvkm_vmm_node_split (struct nvkm_vmm *,struct nvkm_vma *,u64 addr,u64 size);
@@ -173,6 +174,30 @@ void nvkm_vmm_put_locked(struct nvkm_vmm *, struct nvkm_vma *);void nvkm_vmm_unmap_locked (struct nvkm_vmm *,struct nvkm_vma *,bool pfn);void nvkm_vmm_unmap_region (struct nvkm_vmm *,struct nvkm_vma *); +int nvkm_vmm_raw_get (struct nvkm_vmm *vmm,u64 addr,u64 size,u8 refd); +void nvkm_vmm_raw_put (struct nvkm_vmm *vmm,u64 addr,u64 size,u8 refd); +void nvkm_vmm_raw_unmap (struct nvkm_vmm *vmm,u64 addr,u64 size, +bool sparse,u8 refd); +int nvkm_vmm_raw_sparse (struct nvkm_vmm *,u64 addr,u64 size,bool ref); + +static inline bool +nvkm_vmm_in_managed_range (struct nvkm_vmm *vmm,u64 start,u64 size) +{ +u64 p_start = vmm->managed.p.addr; +u64 p_end = p_start + vmm->managed.p.size; +u64 n_start = vmm->managed.n.addr; +u64 n_end = n_start + vmm->managed.n.size; +u64 end = start + size; + +if (start >= p_start && end <= p_end) +return true; + +if (start >= n_start && end <= n_end) +return true; + +return false; +} +#define NVKM_VMM_PFN_ADDR 0xfffffffffffff000ULL#define NVKM_VMM_PFN_ADDR_SHIFT 12#define NVKM_VMM_PFN_APER 0x00000000000000f0ULL
drivers/gpu/drm/nouveau/nvkm/subdev/mmu/vmmgf100.c +9 -7
@@ -287,15 +287,17 @@ gf100_vmm_valid(struct nvkm_vmm *vmm, void *argv, u32 argc,return -EINVAL ; } - ret =nvkm_memory_tags_get (memory, device, tags, - nvkm_ltc_tags_clear, - &map->tags); -if (ret) { -VMM_DEBUG (vmm,"comp %d" , ret); -return ret; +if (!map->no_comp) { + ret =nvkm_memory_tags_get (memory, device, tags, + nvkm_ltc_tags_clear, + &map->tags); +if (ret) { +VMM_DEBUG (vmm,"comp %d" , ret); +return ret; + } } -if (map->tags->mn) { +if (!map->no_comp && map->tags->mn) {u64 tags = map->tags->mn->offset + (map->offset >>17 );if (page->shift ==17 || !gm20x) { map->type |= tags <<44 ;
drivers/gpu/drm/nouveau/nvkm/subdev/mmu/vmmgp100.c +9 -7
@@ -453,15 +453,17 @@ gp100_vmm_valid(struct nvkm_vmm *vmm, void *argv, u32 argc,return -EINVAL ; } - ret =nvkm_memory_tags_get (memory, device, tags, - nvkm_ltc_tags_clear, - &map->tags); -if (ret) { -VMM_DEBUG (vmm,"comp %d" , ret); -return ret; +if (!map->no_comp) { + ret =nvkm_memory_tags_get (memory, device, tags, + nvkm_ltc_tags_clear, + &map->tags); +if (ret) { +VMM_DEBUG (vmm,"comp %d" , ret); +return ret; + } } -if (map->tags->mn) { +if (!map->no_comp && map->tags->mn) { tags = map->tags->mn->offset + (map->offset >>16 ); map->ctag |= ((1ULL << page->shift) >>16 ) <<36 ; map->type |= tags <<36 ;
drivers/gpu/drm/nouveau/nvkm/subdev/mmu/vmmnv50.c +15 -12
@@ -296,19 +296,22 @@ nv50_vmm_valid(struct nvkm_vmm *vmm, void *argv, u32 argc,return -EINVAL ; } - ret =nvkm_memory_tags_get (memory, device, tags,NULL , - &map->tags); -if (ret) { -VMM_DEBUG (vmm,"comp %d" , ret); -return ret; - } +if (!map->no_comp) { + ret =nvkm_memory_tags_get (memory, device, tags,NULL , + &map->tags); +if (ret) { +VMM_DEBUG (vmm,"comp %d" , ret); +return ret; + } -if (map->tags->mn) { -u32 tags = map->tags->mn->offset + (map->offset >>16 ); - map->ctag |= (u64 )comp <<49 ; - map->type |= (u64 )comp <<47 ; - map->type |= (u64 )tags <<49 ; - map->next |= map->ctag; +if (map->tags->mn) { +u32 tags = map->tags->mn->offset + + (map->offset >>16 ); + map->ctag |= (u64 )comp <<49 ; + map->type |= (u64 )comp <<47 ; + map->type |= (u64 )tags <<49 ; + map->next |= map->ctag; + } } }
Signature · 1 line
2.41.0
PATCH v6 12/13 drm/nouveau: implement new VM_BIND uAPI
In reply to Danilo KrummrichThis commit provides the implementation for the new uapi motivated by the Vulkan API. It allows user mode drivers (UMDs) to:
1) Initialize a GPU virtual address (VA) space via the new DRM_IOCTL_NOUVEAU_VM_INIT ioctl for UMDs to specify the portion of VA space managed by the kernel and userspace, respectively.
2) Allocate and free a VA space region as well as bind and unbind memory to the GPUs VA space via the new DRM_IOCTL_NOUVEAU_VM_BIND ioctl. UMDs can request the named operations to be processed either synchronously or asynchronously. It supports DRM syncobjs (incl. timelines) as synchronization mechanism. The management of the GPU VA mappings is implemented with the DRM GPU VA manager.
3) Execute push buffers with the new DRM_IOCTL_NOUVEAU_EXEC ioctl. The execution happens asynchronously. It supports DRM syncobj (incl. timelines) as synchronization mechanism. DRM GEM object locking is handled with drm_exec.
Both, DRM_IOCTL_NOUVEAU_VM_BIND and DRM_IOCTL_NOUVEAU_EXEC, use the DRM GPU scheduler for the asynchronous paths.
--- Documentation/gpu/driver-uapi.rst | 3 + drivers/gpu/drm/nouveau/Kbuild | 3 + drivers/gpu/drm/nouveau/Kconfig | 2 + drivers/gpu/drm/nouveau/nouveau_abi16.c | 24 + drivers/gpu/drm/nouveau/nouveau_abi16.h | 1 + drivers/gpu/drm/nouveau/nouveau_bo.c | 147 +- drivers/gpu/drm/nouveau/nouveau_bo.h | 2 +- drivers/gpu/drm/nouveau/nouveau_drm.c | 27 +- drivers/gpu/drm/nouveau/nouveau_drv.h | 59 +- drivers/gpu/drm/nouveau/nouveau_exec.c | 414 +++++ drivers/gpu/drm/nouveau/nouveau_exec.h | 54 + drivers/gpu/drm/nouveau/nouveau_gem.c | 25 +- drivers/gpu/drm/nouveau/nouveau_mem.h | 5 + drivers/gpu/drm/nouveau/nouveau_prime.c | 2 +- drivers/gpu/drm/nouveau/nouveau_sched.c | 462 ++++++ drivers/gpu/drm/nouveau/nouveau_sched.h | 123 ++ drivers/gpu/drm/nouveau/nouveau_uvmm.c | 1970 +++++++++++++++++++++++ drivers/gpu/drm/nouveau/nouveau_uvmm.h | 107 ++ 18 files changed, 3365 insertions(+), 65 deletions(-) create mode 100644 drivers/gpu/drm/nouveau/nouveau_exec.c create mode 100644 drivers/gpu/drm/nouveau/nouveau_exec.h create mode 100644 drivers/gpu/drm/nouveau/nouveau_sched.c create mode 100644 drivers/gpu/drm/nouveau/nouveau_sched.h create mode 100644 drivers/gpu/drm/nouveau/nouveau_uvmm.c create mode 100644 drivers/gpu/drm/nouveau/nouveau_uvmm.h
Documentation/gpu/driver-uapi.rst +3 -0
@@ -13,4 +13,7 @@ drm/nouveau uAPI VM_BIND / EXEC uAPI ------------------- +.. kernel-doc:: drivers/gpu/drm/nouveau/nouveau_exec.c + :doc: Overview + .. kernel-doc:: include/uapi/drm/nouveau_drm.h
drivers/gpu/drm/nouveau/Kbuild +3 -0
@@ -47,6 +47,9 @@ nouveau-y += nouveau_prime.o nouveau-y += nouveau_sgdma.o nouveau-y += nouveau_ttm.o nouveau-y += nouveau_vmm.o +nouveau-y += nouveau_exec.o +nouveau-y += nouveau_sched.o +nouveau-y += nouveau_uvmm.o # DRM - modesetting nouveau-$(CONFIG_DRM_NOUVEAU_BACKLIGHT) += nouveau_backlight.o
drivers/gpu/drm/nouveau/Kconfig +2 -0
@@ -10,6 +10,8 @@ config DRM_NOUVEAU select DRM_KMS_HELPER select DRM_TTM select DRM_TTM_HELPER + select DRM_EXEC + select DRM_SCHED select I2C select I2C_ALGOBIT select BACKLIGHT_CLASS_DEVICE if DRM_NOUVEAU_BACKLIGHT
drivers/gpu/drm/nouveau/nouveau_abi16.c +24 -0
@@ -35,6 +35,7 @@#include "nouveau_chan.h" #include "nouveau_abi16.h" #include "nouveau_vmm.h" +#include "nouveau_sched.h" static struct nouveau_abi16 *nouveau_abi16 (struct drm_file *file_priv)
@@ -125,6 +126,17 @@ nouveau_abi16_chan_fini(struct nouveau_abi16 *abi16, {struct nouveau_abi16_ntfy *ntfy, *temp; +/* When a client exits without waiting for it's queued up jobs to +* finish it might happen that we fault the channel. This is due to +* drm_file_free() calling drm_gem_release() before the postclose() +* callback. Hence, we can't tear down this scheduler entity before +* uvmm mappings are unmapped. Currently, we can't detect this case. +* +* However, this should be rare and harmless, since the channel isn't +* needed anymore. +*/ +nouveau_sched_entity_fini (&chan->sched_entity); +/* wait for all activity to stop before cleaning up */ if (chan->chan)nouveau_channel_idle (chan->chan);
@@ -261,6 +273,13 @@ nouveau_abi16_ioctl_channel_alloc(ABI16_IOCTL_ARGS)if (!drm->channel)return nouveau_abi16_put (abi16, -ENODEV ); +/* If uvmm wasn't initialized until now disable it completely to prevent +* userspace from mixing up UAPIs. +* +* The client lock is already acquired by nouveau_abi16_get(). +*/ +__nouveau_cli_uvmm_disable (cli); + device = &abi16->device; engine =NV_DEVICE_HOST_RUNLIST_ENGINES_GR ;
@@ -304,6 +323,11 @@ nouveau_abi16_ioctl_channel_alloc(ABI16_IOCTL_ARGS)if (ret) goto done; + ret =nouveau_sched_entity_init (&chan->sched_entity, &drm->sched, + drm->sched_wq); +if (ret) + goto done; + init->channel = chan->chan->chid;if (device->info.family >=NV_DEVICE_INFO_V0_TESLA )
drivers/gpu/drm/nouveau/nouveau_abi16.h +1 -0
@@ -26,6 +26,7 @@ struct nouveau_abi16_chan {struct nouveau_bo *ntfy;struct nouveau_vma *ntfy_vma;struct nvkm_mm heap; +struct nouveau_sched_entity sched_entity; };struct nouveau_abi16 {
drivers/gpu/drm/nouveau/nouveau_bo.c +92 -55
186 lines not on this page · raw message
drivers/gpu/drm/nouveau/nouveau_bo.h +1 -1
@@ -73,7 +73,7 @@ extern struct ttm_device_funcs nouveau_bo_driver;void nouveau_bo_move_init (struct nouveau_drm *);struct nouveau_bo *nouveau_bo_alloc (struct nouveau_cli *,u64 *size,int *align, -u32 domain,u32 tile_mode,u32 tile_flags); +u32 domain,u32 tile_mode,u32 tile_flags,bool internal);int nouveau_bo_init (struct nouveau_bo *,u64 size,int align,u32 domain,struct sg_table *sg,struct dma_resv *robj); int nouveau_bo_new(struct nouveau_cli *, u64 size, int align, u32 domain,
drivers/gpu/drm/nouveau/nouveau_drm.c +25 -2
@@ -68,6 +68,9 @@#include "nouveau_platform.h" #include "nouveau_svm.h" #include "nouveau_dmem.h" +#include "nouveau_exec.h" +#include "nouveau_uvmm.h" +#include "nouveau_sched.h" DECLARE_DYNDBG_CLASSMAP(drm_debug_classes, DD_CLASS_TYPE_DISJOINT_BITS, 0, "DRM_UT_CORE",
@@ -196,6 +199,8 @@ nouveau_cli_fini(struct nouveau_cli *cli)WARN_ON (!list_empty (&cli->worker));usif_client_fini (cli); +nouveau_uvmm_fini (&cli->uvmm); +nouveau_sched_entity_fini (&cli->sched_entity);nouveau_vmm_fini (&cli->svm);nouveau_vmm_fini (&cli->vmm);nvif_mmu_dtor (&cli->mmu);
@@ -301,6 +306,12 @@ nouveau_cli_init(struct nouveau_drm *drm, const char *sname, } cli->mem = &mems[ret]; + + ret =nouveau_sched_entity_init (&cli->sched_entity, &drm->sched, + drm->sched_wq); +if (ret) + goto done; +return 0 ; done:if (ret)
@@ -554,10 +565,14 @@ nouveau_drm_device_init(struct drm_device *dev)nvif_parent_ctor (&nouveau_parent, &drm->parent); drm->master.base.object.parent = &drm->parent; - ret =nouveau_cli_init (drm,"DRM-master" , &drm->master); + ret =nouveau_sched_init (drm);if (ret) goto fail_alloc; + ret =nouveau_cli_init (drm,"DRM-master" , &drm->master); +if (ret) + goto fail_sched; + ret =nouveau_cli_init (drm,"DRM" , &drm->client);if (ret) goto fail_master;
@@ -614,7 +629,6 @@ nouveau_drm_device_init(struct drm_device *dev) }return 0 ; - fail_dispinit:nouveau_display_destroy (dev); fail_dispctor:
@@ -627,6 +641,8 @@ nouveau_drm_device_init(struct drm_device *dev)nouveau_cli_fini (&drm->client); fail_master:nouveau_cli_fini (&drm->master); +fail_sched: +nouveau_sched_fini (drm); fail_alloc:nvif_parent_dtor (&drm->parent);kfree (drm);
@@ -678,6 +694,8 @@ nouveau_drm_device_fini(struct drm_device *dev) }mutex_unlock (&drm->clients_lock); +nouveau_sched_fini (drm); +nouveau_cli_fini (&drm->client);nouveau_cli_fini (&drm->master);nvif_parent_dtor (&drm->parent);
@@ -1179,6 +1197,9 @@ nouveau_ioctls[] = {DRM_IOCTL_DEF_DRV (NOUVEAU_GEM_CPU_PREP , nouveau_gem_ioctl_cpu_prep,DRM_RENDER_ALLOW ),DRM_IOCTL_DEF_DRV (NOUVEAU_GEM_CPU_FINI , nouveau_gem_ioctl_cpu_fini,DRM_RENDER_ALLOW ),DRM_IOCTL_DEF_DRV (NOUVEAU_GEM_INFO , nouveau_gem_ioctl_info,DRM_RENDER_ALLOW ), +DRM_IOCTL_DEF_DRV (NOUVEAU_VM_INIT , nouveau_uvmm_ioctl_vm_init,DRM_RENDER_ALLOW ), +DRM_IOCTL_DEF_DRV (NOUVEAU_VM_BIND , nouveau_uvmm_ioctl_vm_bind,DRM_RENDER_ALLOW ), + DRM_IOCTL_DEF_DRV(NOUVEAU_EXEC, nouveau_exec_ioctl_exec, DRM_RENDER_ALLOW), };long
@@ -1226,6 +1247,8 @@ nouveau_driver_fops = {static struct drm_driver driver_stub = { .driver_features =DRIVER_GEM | +DRIVER_SYNCOBJ |DRIVER_SYNCOBJ_TIMELINE | +DRIVER_GEM_GPUVA |DRIVER_MODESET |DRIVER_RENDER , .open = nouveau_drm_open,
drivers/gpu/drm/nouveau/nouveau_drv.h +57 -2
99 lines not on this page · raw message
drivers/gpu/drm/nouveau/nouveau_exec.c new file +414 -0
414 lines not on this page · raw message
drivers/gpu/drm/nouveau/nouveau_exec.h new file +54 -0
54 lines not on this page · raw message
drivers/gpu/drm/nouveau/nouveau_gem.c +21 -4
@@ -120,7 +120,11 @@ nouveau_gem_object_open(struct drm_gem_object *gem, struct drm_file *file_priv) goto out; } - ret =nouveau_vma_new (nvbo, vmm, &vma); +/* only create a VMA on binding */ +if (!nouveau_cli_uvmm (cli)) + ret =nouveau_vma_new (nvbo, vmm, &vma); +else + ret =0 ;pm_runtime_mark_last_busy (dev);pm_runtime_put_autosuspend (dev); out:
@@ -187,6 +191,9 @@ nouveau_gem_object_close(struct drm_gem_object *gem, struct drm_file *file_priv)if (vmm->vmm.object.oclass <NVIF_CLASS_VMM_NV50 )return ; +if (nouveau_cli_uvmm (cli)) +return ; + ret =ttm_bo_reserve (&nvbo->bo, false, false,NULL );if (ret)return ;
@@ -231,7 +238,7 @@ nouveau_gem_new(struct nouveau_cli *cli, u64 size, int align, uint32_t domain, domain |=NOUVEAU_GEM_DOMAIN_CPU ; nvbo =nouveau_bo_alloc (cli, &size, &align, domain, tile_mode, - tile_flags); + tile_flags, false);if (IS_ERR (nvbo))return PTR_ERR (nvbo);
@@ -279,13 +286,15 @@ nouveau_gem_info(struct drm_file *file_priv, struct drm_gem_object *gem,else rep->domain =NOUVEAU_GEM_DOMAIN_VRAM ; rep->offset = nvbo->offset; -if (vmm->vmm.object.oclass >=NVIF_CLASS_VMM_NV50 ) { +if (vmm->vmm.object.oclass >=NVIF_CLASS_VMM_NV50 && + !nouveau_cli_uvmm (cli)) { vma =nouveau_vma_find (nvbo, vmm);if (!vma)return -EINVAL ; rep->offset = vma->addr; - } + }else + rep->offset =0 ; rep->size = nvbo->bo.base.size; rep->map_handle =drm_vma_node_offset_addr (&nvbo->bo.base.vma_node);
@@ -310,6 +319,11 @@ nouveau_gem_ioctl_new(struct drm_device *dev, void *data,struct nouveau_bo *nvbo =NULL ;int ret =0 ; +/* If uvmm wasn't initialized until now disable it completely to prevent +* userspace from mixing up UAPIs. +*/ +nouveau_cli_uvmm_disable (cli); + ret =nouveau_gem_new (cli, req->info.size, req->align, req->info.domain, req->info.tile_mode, req->info.tile_flags, &nvbo);
@@ -721,6 +735,9 @@ nouveau_gem_ioctl_pushbuf(struct drm_device *dev, void *data,if (unlikely (!abi16))return -ENOMEM ; +if (unlikely (nouveau_cli_uvmm (cli))) +return -ENOSYS ; +list_for_each_entry (temp, &abi16->channels, head) {if (temp->chan->chid == req->channel) { chan = temp->chan;
drivers/gpu/drm/nouveau/nouveau_mem.h +5 -0
@@ -35,4 +35,9 @@ int nouveau_mem_vram(struct ttm_resource *, bool contig, u8 page);int nouveau_mem_host (struct ttm_resource *,struct ttm_tt *);void nouveau_mem_fini (struct nouveau_mem *);int nouveau_mem_map (struct nouveau_mem *,struct nvif_vmm *,struct nvif_vma *); +int +nouveau_mem_map_fixed (struct nouveau_mem *mem, +struct nvif_vmm *vmm, +u8 kind,u64 addr, +u64 offset,u64 range);#endif
drivers/gpu/drm/nouveau/nouveau_prime.c +1 -1
@@ -50,7 +50,7 @@ struct drm_gem_object *nouveau_gem_prime_import_sg_table(struct drm_device *dev,dma_resv_lock (robj,NULL ); nvbo =nouveau_bo_alloc (&drm->client, &size, &align, -NOUVEAU_GEM_DOMAIN_GART ,0 ,0 ); +NOUVEAU_GEM_DOMAIN_GART ,0 ,0 , true);if (IS_ERR (nvbo)) { obj =ERR_CAST (nvbo); goto unlock;
drivers/gpu/drm/nouveau/nouveau_sched.c new file +462 -0
462 lines not on this page · raw message
drivers/gpu/drm/nouveau/nouveau_sched.h new file +123 -0
123 lines not on this page · raw message
drivers/gpu/drm/nouveau/nouveau_uvmm.c new file +1970 -0
1,408 lines not on this page · raw message
drivers/gpu/drm/nouveau/nouveau_uvmm.h new file +107 -0
107 lines not on this page · raw message
The rest of this diff is not shown. View the raw message.
Signature · 1 line
2.41.0
Part of the diff is not shown here. View the raw message.
PATCH v6 13/13 drm/nouveau: debugfs: implement DRM GPU VA debugfs
In reply to Danilo KrummrichProvide the driver indirection iterating over all DRM GPU VA spaces to enable the common 'gpuvas' debugfs file for dumping DRM GPU VA spaces.
--- drivers/gpu/drm/nouveau/nouveau_debugfs.c | 39 +++++++++++++++++++++++ 1 file changed, 39 insertions(+)
drivers/gpu/drm/nouveau/nouveau_debugfs.c +39 -0
@@ -203,6 +203,44 @@ nouveau_debugfs_pstate_open(struct inode *inode, struct file *file)return single_open (file, nouveau_debugfs_pstate_get, inode->i_private); } +static void +nouveau_debugfs_gpuva_regions (struct seq_file *m,struct nouveau_uvmm *uvmm) +{ +MA_STATE (mas, &uvmm->region_mt,0 ,0 ); +struct nouveau_uvma_region *reg; + +seq_puts (m," VA regions | start | range | end \n" ); +seq_puts (m,"----------------------------------------------------------------------------\n" ); +mas_for_each (&mas, reg,ULONG_MAX ) +seq_printf (m," | 0x%016llx | 0x%016llx | 0x%016llx\n" , + reg->va.addr, reg->va.range, reg->va.addr + reg->va.range); +} + +static int +nouveau_debugfs_gpuva (struct seq_file *m,void *data) +{ +struct drm_info_node *node = (struct drm_info_node *) m->private; +struct nouveau_drm *drm =nouveau_drm (node->minor->dev); +struct nouveau_cli *cli; + +mutex_lock (&drm->clients_lock); +list_for_each_entry (cli, &drm->clients, head) { +struct nouveau_uvmm *uvmm =nouveau_cli_uvmm (cli); + +if (!uvmm) +continue ; + +nouveau_uvmm_lock (uvmm); +drm_debugfs_gpuva_info (m, &uvmm->umgr); +seq_puts (m,"\n" ); +nouveau_debugfs_gpuva_regions (m, uvmm); +nouveau_uvmm_unlock (uvmm); + } +mutex_unlock (&drm->clients_lock); + +return 0 ; +} +static const struct file_operations nouveau_pstate_fops = { .owner =THIS_MODULE , .open = nouveau_debugfs_pstate_open,
@@ -214,6 +252,7 @@ static const struct file_operations nouveau_pstate_fops = {static struct drm_info_list nouveau_debugfs_list[] = { {"vbios.rom" , nouveau_debugfs_vbios_image,0 ,NULL }, {"strap_peek" , nouveau_debugfs_strap_peek,0 ,NULL }, +DRM_DEBUGFS_GPUVA_INFO (nouveau_debugfs_gpuva,NULL ), };#define NOUVEAU_DEBUGFS_ENTRIES ARRAY_SIZE(nouveau_debugfs_list)
Signature · 1 line
2.41.0