diff options
author | Matthew Brost <matthew.brost@intel.com> | 2023-12-05 21:39:54 +0300 |
---|---|---|
committer | Rodrigo Vivi <rodrigo.vivi@intel.com> | 2023-12-21 19:46:09 +0300 |
commit | eb9702ad29863c1ae41d17d8504c7444f280dfff (patch) | |
tree | 797ec928e909153489844a892caa6c56984e9bd4 /drivers/gpu/drm/xe/xe_vm.c | |
parent | f5783b5026f76083ef4c53f6240619bd5c7bb9a5 (diff) | |
download | linux-eb9702ad29863c1ae41d17d8504c7444f280dfff.tar.xz |
drm/xe: Allow num_batch_buffer / num_binds == 0 in IOCTLs
The idea being out-syncs can signal indicating all previous operations
on the bind queue are complete. An example use case of this would be
support for implementing vkQueueWaitIdle easily.
All in-syncs are waited on before signaling out-syncs. This is
implemented by forming a composite software fence of in-syncs and
installing this fence in the out-syncs and exec queue last fence slot.
The last fence must be added as a dependency for jobs on user exec
queues as it is possible for the last fence to be a composite software
fence (unordered, ioctl with zero bb or binds) rather than hardware
fence (ordered, previous job on queue).
Cc: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Signed-off-by: Matthew Brost <matthew.brost@intel.com>
Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Diffstat (limited to 'drivers/gpu/drm/xe/xe_vm.c')
-rw-r--r-- | drivers/gpu/drm/xe/xe_vm.c | 77 |
1 files changed, 56 insertions, 21 deletions
diff --git a/drivers/gpu/drm/xe/xe_vm.c b/drivers/gpu/drm/xe/xe_vm.c index d1e53905268f..2f3df9ee67c9 100644 --- a/drivers/gpu/drm/xe/xe_vm.c +++ b/drivers/gpu/drm/xe/xe_vm.c @@ -2722,7 +2722,6 @@ static int vm_bind_ioctl_check_args(struct xe_device *xe, return -EINVAL; if (XE_IOCTL_DBG(xe, args->extensions) || - XE_IOCTL_DBG(xe, !args->num_binds) || XE_IOCTL_DBG(xe, args->num_binds > MAX_BINDS)) return -EINVAL; @@ -2837,6 +2836,37 @@ free_bind_ops: return err; } +static int vm_bind_ioctl_signal_fences(struct xe_vm *vm, + struct xe_exec_queue *q, + struct xe_sync_entry *syncs, + int num_syncs) +{ + struct dma_fence *fence; + int i, err = 0; + + fence = xe_sync_in_fence_get(syncs, num_syncs, + to_wait_exec_queue(vm, q), vm); + if (IS_ERR(fence)) + return PTR_ERR(fence); + + for (i = 0; i < num_syncs; i++) + xe_sync_entry_signal(&syncs[i], NULL, fence); + + xe_exec_queue_last_fence_set(to_wait_exec_queue(vm, q), vm, + fence); + + if (xe_vm_sync_mode(vm, q)) { + long timeout = dma_fence_wait(fence, true); + + if (timeout < 0) + err = -EINTR; + } + + dma_fence_put(fence); + + return err; +} + int xe_vm_bind_ioctl(struct drm_device *dev, void *data, struct drm_file *file) { struct xe_device *xe = to_xe_device(dev); @@ -2875,7 +2905,7 @@ int xe_vm_bind_ioctl(struct drm_device *dev, void *data, struct drm_file *file) goto put_exec_queue; } - if (XE_IOCTL_DBG(xe, async != + if (XE_IOCTL_DBG(xe, args->num_binds && async != !!(q->flags & EXEC_QUEUE_FLAG_VM_ASYNC))) { err = -EINVAL; goto put_exec_queue; @@ -2889,7 +2919,7 @@ int xe_vm_bind_ioctl(struct drm_device *dev, void *data, struct drm_file *file) } if (!args->exec_queue_id) { - if (XE_IOCTL_DBG(xe, async != + if (XE_IOCTL_DBG(xe, args->num_binds && async != !!(vm->flags & XE_VM_FLAG_ASYNC_DEFAULT))) { err = -EINVAL; goto put_vm; @@ -2916,16 +2946,18 @@ int xe_vm_bind_ioctl(struct drm_device *dev, void *data, struct drm_file *file) } } - bos = kzalloc(sizeof(*bos) * args->num_binds, GFP_KERNEL); - if (!bos) { - err = -ENOMEM; - goto release_vm_lock; - } + if (args->num_binds) { + bos = kcalloc(args->num_binds, sizeof(*bos), GFP_KERNEL); + if (!bos) { + err = -ENOMEM; + goto release_vm_lock; + } - ops = kzalloc(sizeof(*ops) * args->num_binds, GFP_KERNEL); - if (!ops) { - err = -ENOMEM; - goto release_vm_lock; + ops = kcalloc(args->num_binds, sizeof(*ops), GFP_KERNEL); + if (!ops) { + err = -ENOMEM; + goto release_vm_lock; + } } for (i = 0; i < args->num_binds; ++i) { @@ -2995,12 +3027,19 @@ int xe_vm_bind_ioctl(struct drm_device *dev, void *data, struct drm_file *file) for (num_syncs = 0; num_syncs < args->num_syncs; num_syncs++) { err = xe_sync_entry_parse(xe, xef, &syncs[num_syncs], &syncs_user[num_syncs], - xe_vm_in_lr_mode(vm) ? - SYNC_PARSE_FLAG_LR_MODE : 0); + (xe_vm_in_lr_mode(vm) ? + SYNC_PARSE_FLAG_LR_MODE : 0) | + (!args->num_binds ? + SYNC_PARSE_FLAG_DISALLOW_USER_FENCE : 0)); if (err) goto free_syncs; } + if (!args->num_binds) { + err = -ENODATA; + goto free_syncs; + } + for (i = 0; i < args->num_binds; ++i) { u64 range = bind_ops[i].range; u64 addr = bind_ops[i].addr; @@ -3058,12 +3097,8 @@ int xe_vm_bind_ioctl(struct drm_device *dev, void *data, struct drm_file *file) unwind_ops: vm_bind_ioctl_ops_unwind(vm, ops, args->num_binds); free_syncs: - for (i = 0; err == -ENODATA && i < num_syncs; i++) { - struct dma_fence *fence = - xe_exec_queue_last_fence_get(to_wait_exec_queue(vm, q), vm); - - xe_sync_entry_signal(&syncs[i], NULL, fence); - } + if (err == -ENODATA) + err = vm_bind_ioctl_signal_fences(vm, q, syncs, num_syncs); while (num_syncs--) xe_sync_entry_cleanup(&syncs[num_syncs]); @@ -3083,7 +3118,7 @@ free_objs: kfree(ops); if (args->num_binds > 1) kfree(bind_ops); - return err == -ENODATA ? 0 : err; + return err; } /** |