diff options
author | Lionel Landwerlin <lionel.g.landwerlin@intel.com> | 2020-08-04 11:59:54 +0300 |
---|---|---|
committer | Rodrigo Vivi <rodrigo.vivi@intel.com> | 2020-08-17 16:16:51 -0400 |
commit | 13149e8bafc4657254831ba6c16ed8780aa64a06 (patch) | |
tree | 4f0822c794b8dcb7287815eb4a07db42ae6d6cbb /drivers/gpu/drm/i915 | |
parent | cda9edd02425d7902714c60a6f6e31881d2f2741 (diff) |
drm/i915: add syncobj timeline support
Introduces a new parameters to execbuf so that we can specify syncobj
handles as well as timeline points.
v2: Reuse i915_user_extension_fn
v3: Check that the chained extension is only present once (Chris)
v4: Check that dma_fence_chain_find_seqno returns a non NULL fence (Lionel)
v5: Use BIT_ULL (Chris)
v6: Fix issue with already signaled timeline points,
dma_fence_chain_find_seqno() setting fence to NULL (Chris)
v7: Report ENOENT with invalid syncobj handle (Lionel)
v8: Check for out of order timeline point insertion (Chris)
v9: After explanations on
https://lists.freedesktop.org/archives/dri-devel/2019-August/229287.html
drop the ordering check from v8 (Lionel)
v10: Set first extension enum item to 1 (Jason)
v11: Rebase
v12: Allow multiple extension nodes of timeline syncobj (Chris)
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Co-authored-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch> (v11)
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20200804085954.350343-3-lionel.g.landwerlin@intel.com
Link: https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/2901
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Diffstat (limited to 'drivers/gpu/drm/i915')
-rw-r--r-- | drivers/gpu/drm/i915/gem/i915_gem_execbuffer.c | 312 | ||||
-rw-r--r-- | drivers/gpu/drm/i915/i915_drv.c | 3 | ||||
-rw-r--r-- | drivers/gpu/drm/i915/i915_getparam.c | 1 |
3 files changed, 243 insertions, 73 deletions
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/gem/i915_gem_execbuffer.c index 09d2f955b11e..02b1630f513e 100644 --- a/drivers/gpu/drm/i915/gem/i915_gem_execbuffer.c +++ b/drivers/gpu/drm/i915/gem/i915_gem_execbuffer.c @@ -223,6 +223,13 @@ struct eb_vma_array { * the batchbuffer in trusted mode, otherwise the ioctl is rejected. */ +struct eb_fence { + struct drm_syncobj *syncobj; /* Use with ptr_mask_bits() */ + struct dma_fence *dma_fence; + u64 value; + struct dma_fence_chain *chain_fence; +}; + struct i915_execbuffer { struct drm_i915_private *i915; /** i915 backpointer */ struct drm_file *file; /** per-file lookup tables and limits */ @@ -283,12 +290,8 @@ struct i915_execbuffer { struct hlist_head *buckets; /** ht for relocation handles */ struct eb_vma_array *array; - struct i915_eb_fence { - struct drm_syncobj *syncobj; /* Use with ptr_mask_bits() */ - } *fences; - u32 n_fences; - - u64 extension_flags; /** Available extensions parameters */ + struct eb_fence *fences; + unsigned long num_fences; }; static inline bool eb_use_cmdparser(const struct i915_execbuffer *eb) @@ -2210,77 +2213,222 @@ eb_pin_engine(struct i915_execbuffer *eb, } static void -__free_fence_array(struct i915_eb_fence *fences, unsigned int n) +__free_fence_array(struct eb_fence *fences, unsigned int n) { - while (n--) + while (n--) { drm_syncobj_put(ptr_mask_bits(fences[n].syncobj, 2)); + dma_fence_put(fences[n].dma_fence); + kfree(fences[n].chain_fence); + } kvfree(fences); } static int -get_fence_array(struct drm_i915_gem_execbuffer2 *args, - struct i915_execbuffer *eb) +add_timeline_fence_array(struct i915_execbuffer *eb, + const struct drm_i915_gem_execbuffer_ext_timeline_fences *timeline_fences) { - const unsigned long nfences = args->num_cliprects; - struct drm_i915_gem_exec_fence __user *user; - struct i915_eb_fence *fences; - unsigned long n; - int err; + struct drm_i915_gem_exec_fence __user *user_fences; + u64 __user *user_values; + struct eb_fence *f; + u64 nfences; + int err = 0; - if (!(args->flags & I915_EXEC_FENCE_ARRAY)) + nfences = timeline_fences->fence_count; + if (!nfences) return 0; /* Check multiplication overflow for access_ok() and kvmalloc_array() */ BUILD_BUG_ON(sizeof(size_t) > sizeof(unsigned long)); if (nfences > min_t(unsigned long, - ULONG_MAX / sizeof(*user), - SIZE_MAX / sizeof(*fences))) + ULONG_MAX / sizeof(*user_fences), + SIZE_MAX / sizeof(*f)) - eb->num_fences) return -EINVAL; - user = u64_to_user_ptr(args->cliprects_ptr); - if (!access_ok(user, nfences * sizeof(*user))) + user_fences = u64_to_user_ptr(timeline_fences->handles_ptr); + if (!access_ok(user_fences, nfences * sizeof(*user_fences))) + return -EFAULT; + + user_values = u64_to_user_ptr(timeline_fences->values_ptr); + if (!access_ok(user_values, nfences * sizeof(*user_values))) return -EFAULT; - fences = kvmalloc_array(nfences, sizeof(*fences), - __GFP_NOWARN | GFP_KERNEL); - if (!fences) + f = krealloc(eb->fences, + (eb->num_fences + nfences) * sizeof(*f), + __GFP_NOWARN | GFP_KERNEL); + if (!f) return -ENOMEM; - for (n = 0; n < nfences; n++) { - struct drm_i915_gem_exec_fence fence; + eb->fences = f; + f += eb->num_fences; + + BUILD_BUG_ON(~(ARCH_KMALLOC_MINALIGN - 1) & + ~__I915_EXEC_FENCE_UNKNOWN_FLAGS); + + while (nfences--) { + struct drm_i915_gem_exec_fence user_fence; struct drm_syncobj *syncobj; + struct dma_fence *fence = NULL; + u64 point; + + if (__copy_from_user(&user_fence, + user_fences++, + sizeof(user_fence))) + return -EFAULT; + + if (user_fence.flags & __I915_EXEC_FENCE_UNKNOWN_FLAGS) + return -EINVAL; + + if (__get_user(point, user_values++)) + return -EFAULT; + + syncobj = drm_syncobj_find(eb->file, user_fence.handle); + if (!syncobj) { + DRM_DEBUG("Invalid syncobj handle provided\n"); + return -ENOENT; + } + + fence = drm_syncobj_fence_get(syncobj); - if (__copy_from_user(&fence, user++, sizeof(fence))) { - err = -EFAULT; - goto err; + if (!fence && user_fence.flags && + !(user_fence.flags & I915_EXEC_FENCE_SIGNAL)) { + DRM_DEBUG("Syncobj handle has no fence\n"); + drm_syncobj_put(syncobj); + return -EINVAL; } - if (fence.flags & __I915_EXEC_FENCE_UNKNOWN_FLAGS) { - err = -EINVAL; - goto err; + if (fence) + err = dma_fence_chain_find_seqno(&fence, point); + + if (err && !(user_fence.flags & I915_EXEC_FENCE_SIGNAL)) { + DRM_DEBUG("Syncobj handle missing requested point %llu\n", point); + drm_syncobj_put(syncobj); + return err; + } + + /* + * A point might have been signaled already and + * garbage collected from the timeline. In this case + * just ignore the point and carry on. + */ + if (!fence && !(user_fence.flags & I915_EXEC_FENCE_SIGNAL)) { + drm_syncobj_put(syncobj); + continue; + } + + /* + * For timeline syncobjs we need to preallocate chains for + * later signaling. + */ + if (point != 0 && user_fence.flags & I915_EXEC_FENCE_SIGNAL) { + /* + * Waiting and signaling the same point (when point != + * 0) would break the timeline. + */ + if (user_fence.flags & I915_EXEC_FENCE_WAIT) { + DRM_DEBUG("Trying to wait & signal the same timeline point.\n"); + dma_fence_put(fence); + drm_syncobj_put(syncobj); + return -EINVAL; + } + + f->chain_fence = + kmalloc(sizeof(*f->chain_fence), + GFP_KERNEL); + if (!f->chain_fence) { + drm_syncobj_put(syncobj); + dma_fence_put(fence); + return -ENOMEM; + } + } else { + f->chain_fence = NULL; } - syncobj = drm_syncobj_find(eb->file, fence.handle); + f->syncobj = ptr_pack_bits(syncobj, user_fence.flags, 2); + f->dma_fence = fence; + f->value = point; + f++; + eb->num_fences++; + } + + return 0; +} + +static int add_fence_array(struct i915_execbuffer *eb) +{ + struct drm_i915_gem_execbuffer2 *args = eb->args; + struct drm_i915_gem_exec_fence __user *user; + unsigned long num_fences = args->num_cliprects; + struct eb_fence *f; + + if (!(args->flags & I915_EXEC_FENCE_ARRAY)) + return 0; + + if (!num_fences) + return 0; + + /* Check multiplication overflow for access_ok() and kvmalloc_array() */ + BUILD_BUG_ON(sizeof(size_t) > sizeof(unsigned long)); + if (num_fences > min_t(unsigned long, + ULONG_MAX / sizeof(*user), + SIZE_MAX / sizeof(*f) - eb->num_fences)) + return -EINVAL; + + user = u64_to_user_ptr(args->cliprects_ptr); + if (!access_ok(user, num_fences * sizeof(*user))) + return -EFAULT; + + f = krealloc(eb->fences, + (eb->num_fences + num_fences) * sizeof(*f), + __GFP_NOWARN | GFP_KERNEL); + if (!f) + return -ENOMEM; + + eb->fences = f; + f += eb->num_fences; + while (num_fences--) { + struct drm_i915_gem_exec_fence user_fence; + struct drm_syncobj *syncobj; + struct dma_fence *fence = NULL; + + if (__copy_from_user(&user_fence, user++, sizeof(user_fence))) + return -EFAULT; + + if (user_fence.flags & __I915_EXEC_FENCE_UNKNOWN_FLAGS) + return -EINVAL; + + syncobj = drm_syncobj_find(eb->file, user_fence.handle); if (!syncobj) { DRM_DEBUG("Invalid syncobj handle provided\n"); - err = -ENOENT; - goto err; + return -ENOENT; + } + + if (user_fence.flags & I915_EXEC_FENCE_WAIT) { + fence = drm_syncobj_fence_get(syncobj); + if (!fence) { + DRM_DEBUG("Syncobj handle has no fence\n"); + drm_syncobj_put(syncobj); + return -EINVAL; + } } BUILD_BUG_ON(~(ARCH_KMALLOC_MINALIGN - 1) & ~__I915_EXEC_FENCE_UNKNOWN_FLAGS); - fences[n].syncobj = ptr_pack_bits(syncobj, fence.flags, 2); + f->syncobj = ptr_pack_bits(syncobj, user_fence.flags, 2); + f->dma_fence = fence; + f->value = 0; + f->chain_fence = NULL; + f++; + eb->num_fences++; } - eb->fences = fences; - eb->n_fences = nfences; - return 0; +} -err: - __free_fence_array(fences, n); - return err; +static void put_fence_array(struct eb_fence *fences, int num_fences) +{ + if (fences) + __free_fence_array(fences, num_fences); } static int @@ -2289,21 +2437,17 @@ await_fence_array(struct i915_execbuffer *eb) unsigned int n; int err; - for (n = 0; n < eb->n_fences; n++) { + for (n = 0; n < eb->num_fences; n++) { struct drm_syncobj *syncobj; - struct dma_fence *fence; unsigned int flags; syncobj = ptr_unpack_bits(eb->fences[n].syncobj, &flags, 2); - if (!(flags & I915_EXEC_FENCE_WAIT)) - continue; - fence = drm_syncobj_fence_get(syncobj); - if (!fence) - return -EINVAL; + if (!eb->fences[n].dma_fence) + continue; - err = i915_request_await_dma_fence(eb->request, fence); - dma_fence_put(fence); + err = i915_request_await_dma_fence(eb->request, + eb->fences[n].dma_fence); if (err < 0) return err; } @@ -2311,13 +2455,12 @@ await_fence_array(struct i915_execbuffer *eb) return 0; } -static void -signal_fence_array(struct i915_execbuffer *eb) +static void signal_fence_array(const struct i915_execbuffer *eb) { struct dma_fence * const fence = &eb->request->fence; unsigned int n; - for (n = 0; n < eb->n_fences; n++) { + for (n = 0; n < eb->num_fences; n++) { struct drm_syncobj *syncobj; unsigned int flags; @@ -2325,10 +2468,34 @@ signal_fence_array(struct i915_execbuffer *eb) if (!(flags & I915_EXEC_FENCE_SIGNAL)) continue; - drm_syncobj_replace_fence(syncobj, fence); + if (eb->fences[n].chain_fence) { + drm_syncobj_add_point(syncobj, + eb->fences[n].chain_fence, + fence, + eb->fences[n].value); + /* + * The chain's ownership is transferred to the + * timeline. + */ + eb->fences[n].chain_fence = NULL; + } else { + drm_syncobj_replace_fence(syncobj, fence); + } } } +static int +parse_timeline_fences(struct i915_user_extension __user *ext, void *data) +{ + struct i915_execbuffer *eb = data; + struct drm_i915_gem_execbuffer_ext_timeline_fences timeline_fences; + + if (copy_from_user(&timeline_fences, ext, sizeof(timeline_fences))) + return -EFAULT; + + return add_timeline_fence_array(eb, &timeline_fences); +} + static void retire_requests(struct intel_timeline *tl, struct i915_request *end) { struct i915_request *rq, *rn; @@ -2371,14 +2538,13 @@ static void eb_request_add(struct i915_execbuffer *eb) } static const i915_user_extension_fn execbuf_extensions[] = { + [DRM_I915_GEM_EXECBUFFER_EXT_TIMELINE_FENCES] = parse_timeline_fences, }; static int parse_execbuf2_extensions(struct drm_i915_gem_execbuffer2 *args, struct i915_execbuffer *eb) { - eb->extension_flags = 0; - if (!(args->flags & I915_EXEC_USE_EXTENSIONS)) return 0; @@ -2432,7 +2598,7 @@ i915_gem_do_execbuffer(struct drm_device *dev, eb.trampoline = NULL; eb.fences = NULL; - eb.n_fences = 0; + eb.num_fences = 0; eb.batch_flags = 0; if (args->flags & I915_EXEC_SECURE) { @@ -2451,14 +2617,24 @@ i915_gem_do_execbuffer(struct drm_device *dev, if (args->flags & I915_EXEC_IS_PINNED) eb.batch_flags |= I915_DISPATCH_PINNED; + err = parse_execbuf2_extensions(args, &eb); + if (err) + goto err_ext; + + err = add_fence_array(&eb); + if (err) + goto err_ext; + #define IN_FENCES (I915_EXEC_FENCE_IN | I915_EXEC_FENCE_SUBMIT) if (args->flags & IN_FENCES) { if ((args->flags & IN_FENCES) == IN_FENCES) return -EINVAL; in_fence = sync_file_get_fence(lower_32_bits(args->rsvd2)); - if (!in_fence) - return -EINVAL; + if (!in_fence) { + err = -EINVAL; + goto err_ext; + } } #undef IN_FENCES @@ -2470,17 +2646,9 @@ i915_gem_do_execbuffer(struct drm_device *dev, } } - err = parse_execbuf2_extensions(args, &eb); - if (err) - goto err_out_fence; - - err = get_fence_array(args, &eb); - if (err) - goto err_arr_fence; - err = eb_create(&eb); if (err) - goto err_arr_fence; + goto err_out_fence; GEM_BUG_ON(!eb.lut_size); @@ -2576,7 +2744,7 @@ i915_gem_do_execbuffer(struct drm_device *dev, goto err_request; } - if (eb.n_fences) { + if (eb.fences) { err = await_fence_array(&eb); if (err) goto err_request; @@ -2608,7 +2776,7 @@ err_request: i915_request_get(eb.request); eb_request_add(&eb); - if (eb.n_fences) + if (eb.fences) signal_fence_array(&eb); if (out_fence) { @@ -2637,13 +2805,13 @@ err_context: i915_gem_context_put(eb.gem_context); err_destroy: eb_destroy(&eb); -err_arr_fence: - __free_fence_array(eb.fences, eb.n_fences); err_out_fence: if (out_fence_fd != -1) put_unused_fd(out_fence_fd); err_in_fence: dma_fence_put(in_fence); +err_ext: + put_fence_array(eb.fences, eb.num_fences); return err; } diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c index 5fd5af4bc855..2d10f6a2c042 100644 --- a/drivers/gpu/drm/i915/i915_drv.c +++ b/drivers/gpu/drm/i915/i915_drv.c @@ -1846,7 +1846,8 @@ static struct drm_driver driver = { */ .driver_features = DRIVER_GEM | - DRIVER_RENDER | DRIVER_MODESET | DRIVER_ATOMIC | DRIVER_SYNCOBJ, + DRIVER_RENDER | DRIVER_MODESET | DRIVER_ATOMIC | DRIVER_SYNCOBJ | + DRIVER_SYNCOBJ_TIMELINE, .release = i915_driver_release, .open = i915_driver_open, .lastclose = i915_driver_lastclose, diff --git a/drivers/gpu/drm/i915/i915_getparam.c b/drivers/gpu/drm/i915/i915_getparam.c index 421613219ae9..f96032c60a12 100644 --- a/drivers/gpu/drm/i915/i915_getparam.c +++ b/drivers/gpu/drm/i915/i915_getparam.c @@ -132,6 +132,7 @@ int i915_getparam_ioctl(struct drm_device *dev, void *data, case I915_PARAM_HAS_EXEC_BATCH_FIRST: case I915_PARAM_HAS_EXEC_FENCE_ARRAY: case I915_PARAM_HAS_EXEC_SUBMIT_FENCE: + case I915_PARAM_HAS_EXEC_TIMELINE_FENCES: /* For the time being all of these are always true; * if some supported hardware does not have one of these * features this value needs to be provided from |