Re: [PATCH v2 5/5] drm/vc4: Expose dma-buf fences for V3D rendering.

From: Daniel Vetter
Date: Thu Apr 13 2017 - 03:13:22 EST

On Wed, Apr 12, 2017 at 12:12:02PM -0700, Eric Anholt wrote:
> This is needed for proper synchronization with display on another DRM
> device (pl111 or tinydrm) with buffers produced by vc4 V3D. Fixes the
> new igt vc4_dmabuf_poll testcase, and rendering of one of the glmark2
> desktop tests on pl111+vc4.
>
> This doesn't yet introduce waits on other device's fences before vc4's
> rendering/display, because I don't have testcases for them.
>
> v2: Reuse dma_fence_free(), retitle commit message to clarify that
> it's not a full dma-buf fencing implementation yet.
>
> Signed-off-by: Eric Anholt <eric@xxxxxxxxxx>

Double-checked a few things in your ww_mutex scheme, seems are correct.
And testing with CONFIG_DEBUG_WW_MUTEX_SLOWPATH should catch any kind of
fumbles in your error paths. I didnt do a full review, so just

Acked-by: Daniel Vetter <daniel.vetter@xxxxxxxx>

> ---
> drivers/gpu/drm/vc4/Makefile | 1 +
> drivers/gpu/drm/vc4/vc4_bo.c | 37 ++++++++++-
> drivers/gpu/drm/vc4/vc4_drv.c | 3 +-
> drivers/gpu/drm/vc4/vc4_drv.h | 30 +++++++++
> drivers/gpu/drm/vc4/vc4_fence.c | 56 +++++++++++++++++
> drivers/gpu/drm/vc4/vc4_gem.c | 136 +++++++++++++++++++++++++++++++++++++++-
> drivers/gpu/drm/vc4/vc4_irq.c | 4 ++
> 7 files changed, 262 insertions(+), 5 deletions(-)
> create mode 100644 drivers/gpu/drm/vc4/vc4_fence.c
>
> diff --git a/drivers/gpu/drm/vc4/Makefile b/drivers/gpu/drm/vc4/Makefile
> index 61f45d122bd0..ab687fba4916 100644
> --- a/drivers/gpu/drm/vc4/Makefile
> +++ b/drivers/gpu/drm/vc4/Makefile
> @@ -9,6 +9,7 @@ vc4-y := \
> vc4_drv.o \
> vc4_dpi.o \
> vc4_dsi.o \
> + vc4_fence.o \
> vc4_kms.o \
> vc4_gem.o \
> vc4_hdmi.o \
> diff --git a/drivers/gpu/drm/vc4/vc4_bo.c b/drivers/gpu/drm/vc4/vc4_bo.c
> index af29432a6471..80b2f9e55c5c 100644
> --- a/drivers/gpu/drm/vc4/vc4_bo.c
> +++ b/drivers/gpu/drm/vc4/vc4_bo.c
> @@ -19,6 +19,8 @@
> * rendering can return quickly.
> */
>
> +#include <linux/dma-buf.h>
> +
> #include "vc4_drv.h"
> #include "uapi/drm/vc4_drm.h"
>
> @@ -88,6 +90,10 @@ static void vc4_bo_destroy(struct vc4_bo *bo)
>
> vc4->bo_stats.num_allocated--;
> vc4->bo_stats.size_allocated -= obj->size;
> +
> + if (bo->resv == &bo->_resv)
> + reservation_object_fini(bo->resv);
> +
> drm_gem_cma_free_object(obj);
> }
>
> @@ -244,8 +250,12 @@ struct vc4_bo *vc4_bo_create(struct drm_device *dev, size_t unaligned_size,
> return ERR_PTR(-ENOMEM);
> }
> }
> + bo = to_vc4_bo(&cma_obj->base);
>
> - return to_vc4_bo(&cma_obj->base);
> + bo->resv = &bo->_resv;
> + reservation_object_init(bo->resv);
> +
> + return bo;
> }
>
> int vc4_dumb_create(struct drm_file *file_priv,
> @@ -369,6 +379,13 @@ static void vc4_bo_cache_time_timer(unsigned long data)
> schedule_work(&vc4->bo_cache.time_work);
> }
>
> +struct reservation_object *vc4_prime_res_obj(struct drm_gem_object *obj)
> +{
> + struct vc4_bo *bo = to_vc4_bo(obj);
> +
> + return bo->resv;
> +}
> +
> struct dma_buf *
> vc4_prime_export(struct drm_device *dev, struct drm_gem_object *obj, int flags)
> {
> @@ -440,6 +457,24 @@ void *vc4_prime_vmap(struct drm_gem_object *obj)
> return drm_gem_cma_prime_vmap(obj);
> }
>
> +struct drm_gem_object *
> +vc4_prime_import_sg_table(struct drm_device *dev,
> + struct dma_buf_attachment *attach,
> + struct sg_table *sgt)
> +{
> + struct drm_gem_object *obj;
> + struct vc4_bo *bo;
> +
> + obj = drm_gem_cma_prime_import_sg_table(dev, attach, sgt);
> + if (IS_ERR(obj))
> + return obj;
> +
> + bo = to_vc4_bo(obj);
> + bo->resv = attach->dmabuf->resv;
> +
> + return obj;
> +}
> +
> int vc4_create_bo_ioctl(struct drm_device *dev, void *data,
> struct drm_file *file_priv)
> {
> diff --git a/drivers/gpu/drm/vc4/vc4_drv.c b/drivers/gpu/drm/vc4/vc4_drv.c
> index 61e674baf3a6..92fb9a41fe7c 100644
> --- a/drivers/gpu/drm/vc4/vc4_drv.c
> +++ b/drivers/gpu/drm/vc4/vc4_drv.c
> @@ -168,8 +168,9 @@ static struct drm_driver vc4_drm_driver = {
> .prime_fd_to_handle = drm_gem_prime_fd_to_handle,
> .gem_prime_import = drm_gem_prime_import,
> .gem_prime_export = vc4_prime_export,
> + .gem_prime_res_obj = vc4_prime_res_obj,
> .gem_prime_get_sg_table = drm_gem_cma_prime_get_sg_table,
> - .gem_prime_import_sg_table = drm_gem_cma_prime_import_sg_table,
> + .gem_prime_import_sg_table = vc4_prime_import_sg_table,
> .gem_prime_vmap = vc4_prime_vmap,
> .gem_prime_vunmap = drm_gem_cma_prime_vunmap,
> .gem_prime_mmap = vc4_prime_mmap,
> diff --git a/drivers/gpu/drm/vc4/vc4_drv.h b/drivers/gpu/drm/vc4/vc4_drv.h
> index dea304966e65..08d5c2213c80 100644
> --- a/drivers/gpu/drm/vc4/vc4_drv.h
> +++ b/drivers/gpu/drm/vc4/vc4_drv.h
> @@ -8,7 +8,9 @@
>
> #include "drmP.h"
> #include "drm_gem_cma_helper.h"
> +#include "drm_gem_cma_helper.h"
>
> +#include <linux/reservation.h>
> #include <drm/drm_encoder.h>
>
> struct vc4_dev {
> @@ -56,6 +58,8 @@ struct vc4_dev {
> /* Protects bo_cache and the BO stats. */
> struct mutex bo_lock;
>
> + uint64_t dma_fence_context;
> +
> /* Sequence number for the last job queued in bin_job_list.
> * Starts at 0 (no jobs emitted).
> */
> @@ -150,6 +154,10 @@ struct vc4_bo {
> * DRM_IOCTL_VC4_CREATE_SHADER_BO.
> */
> struct vc4_validated_shader_info *validated_shader;
> +
> + /* normally (resv == &_resv) except for imported bo's */
> + struct reservation_object *resv;
> + struct reservation_object _resv;
> };
>
> static inline struct vc4_bo *
> @@ -158,6 +166,19 @@ to_vc4_bo(struct drm_gem_object *bo)
> return (struct vc4_bo *)bo;
> }
>
> +struct vc4_fence {
> + struct dma_fence base;
> + struct drm_device *dev;
> + /* vc4 seqno for signaled() test */
> + uint64_t seqno;
> +};
> +
> +static inline struct vc4_fence *
> +to_vc4_fence(struct dma_fence *fence)
> +{
> + return (struct vc4_fence *)fence;
> +}
> +
> struct vc4_seqno_cb {
> struct work_struct work;
> uint64_t seqno;
> @@ -231,6 +252,8 @@ struct vc4_exec_info {
> /* Latest write_seqno of any BO that binning depends on. */
> uint64_t bin_dep_seqno;
>
> + struct dma_fence *fence;
> +
> /* Last current addresses the hardware was processing when the
> * hangcheck timer checked on us.
> */
> @@ -437,7 +460,11 @@ int vc4_mmap_bo_ioctl(struct drm_device *dev, void *data,
> int vc4_get_hang_state_ioctl(struct drm_device *dev, void *data,
> struct drm_file *file_priv);
> int vc4_mmap(struct file *filp, struct vm_area_struct *vma);
> +struct reservation_object *vc4_prime_res_obj(struct drm_gem_object *obj);
> int vc4_prime_mmap(struct drm_gem_object *obj, struct vm_area_struct *vma);
> +struct drm_gem_object *vc4_prime_import_sg_table(struct drm_device *dev,
> + struct dma_buf_attachment *attach,
> + struct sg_table *sgt);
> void *vc4_prime_vmap(struct drm_gem_object *obj);
> void vc4_bo_cache_init(struct drm_device *dev);
> void vc4_bo_cache_destroy(struct drm_device *dev);
> @@ -469,6 +496,9 @@ int vc4_dpi_debugfs_regs(struct seq_file *m, void *unused);
> extern struct platform_driver vc4_dsi_driver;
> int vc4_dsi_debugfs_regs(struct seq_file *m, void *unused);
>
> +/* vc4_fence.c */
> +extern const struct dma_fence_ops vc4_fence_ops;
> +
> /* vc4_gem.c */
> void vc4_gem_init(struct drm_device *dev);
> void vc4_gem_destroy(struct drm_device *dev);
> diff --git a/drivers/gpu/drm/vc4/vc4_fence.c b/drivers/gpu/drm/vc4/vc4_fence.c
> new file mode 100644
> index 000000000000..dbf5a5a5d5f5
> --- /dev/null
> +++ b/drivers/gpu/drm/vc4/vc4_fence.c
> @@ -0,0 +1,56 @@
> +/*
> + * Copyright © 2017 Broadcom
> + *
> + * 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 (including the next
> + * paragraph) 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 AUTHORS OR COPYRIGHT HOLDERS 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 "vc4_drv.h"
> +
> +static const char *vc4_fence_get_driver_name(struct dma_fence *fence)
> +{
> + return "vc4";
> +}
> +
> +static const char *vc4_fence_get_timeline_name(struct dma_fence *fence)
> +{
> + return "vc4-v3d";
> +}
> +
> +static bool vc4_fence_enable_signaling(struct dma_fence *fence)
> +{
> + return true;
> +}
> +
> +static bool vc4_fence_signaled(struct dma_fence *fence)
> +{
> + struct vc4_fence *f = to_vc4_fence(fence);
> + struct vc4_dev *vc4 = to_vc4_dev(f->dev);
> +
> + return vc4->finished_seqno >= f->seqno;
> +}
> +
> +const struct dma_fence_ops vc4_fence_ops = {
> + .get_driver_name = vc4_fence_get_driver_name,
> + .get_timeline_name = vc4_fence_get_timeline_name,
> + .enable_signaling = vc4_fence_enable_signaling,
> + .signaled = vc4_fence_signaled,
> + .wait = dma_fence_default_wait,
> + .release = dma_fence_free,
> +};
> diff --git a/drivers/gpu/drm/vc4/vc4_gem.c b/drivers/gpu/drm/vc4/vc4_gem.c
> index e9c381c42139..a1a01044263c 100644
> --- a/drivers/gpu/drm/vc4/vc4_gem.c
> +++ b/drivers/gpu/drm/vc4/vc4_gem.c
> @@ -463,6 +463,8 @@ vc4_update_bo_seqnos(struct vc4_exec_info *exec, uint64_t seqno)
> for (i = 0; i < exec->bo_count; i++) {
> bo = to_vc4_bo(&exec->bo[i]->base);
> bo->seqno = seqno;
> +
> + reservation_object_add_shared_fence(bo->resv, exec->fence);
> }
>
> list_for_each_entry(bo, &exec->unref_list, unref_head) {
> @@ -472,7 +474,103 @@ vc4_update_bo_seqnos(struct vc4_exec_info *exec, uint64_t seqno)
> for (i = 0; i < exec->rcl_write_bo_count; i++) {
> bo = to_vc4_bo(&exec->rcl_write_bo[i]->base);
> bo->write_seqno = seqno;
> +
> + reservation_object_add_excl_fence(bo->resv, exec->fence);
> + }
> +}
> +
> +static void
> +vc4_unlock_bo_reservations(struct drm_device *dev,
> + struct vc4_exec_info *exec,
> + struct ww_acquire_ctx *acquire_ctx)
> +{
> + int i;
> +
> + for (i = 0; i < exec->bo_count; i++) {
> + struct vc4_bo *bo = to_vc4_bo(&exec->bo[i]->base);
> +
> + ww_mutex_unlock(&bo->resv->lock);
> }
> +
> + ww_acquire_fini(acquire_ctx);
> +}
> +
> +/* Takes the reservation lock on all the BOs being referenced, so that
> + * at queue submit time we can update the reservations.
> + *
> + * We don't lock the RCL the tile alloc/state BOs, or overflow memory
> + * (all of which are on exec->unref_list). They're entirely private
> + * to vc4, so we don't attach dma-buf fences to them.
> + */
> +static int
> +vc4_lock_bo_reservations(struct drm_device *dev,
> + struct vc4_exec_info *exec,
> + struct ww_acquire_ctx *acquire_ctx)
> +{
> + int contended_lock = -1;
> + int i, ret;
> + struct vc4_bo *bo;
> +
> + ww_acquire_init(acquire_ctx, &reservation_ww_class);
> +
> +retry:
> + if (contended_lock != -1) {
> + bo = to_vc4_bo(&exec->bo[contended_lock]->base);
> + ret = ww_mutex_lock_slow_interruptible(&bo->resv->lock,
> + acquire_ctx);
> + if (ret) {
> + ww_acquire_done(acquire_ctx);
> + return ret;
> + }
> + }
> +
> + for (i = 0; i < exec->bo_count; i++) {
> + if (i == contended_lock)
> + continue;
> +
> + bo = to_vc4_bo(&exec->bo[i]->base);
> +
> + ret = ww_mutex_lock_interruptible(&bo->resv->lock, acquire_ctx);
> + if (ret) {
> + int j;
> +
> + for (j = 0; j < i; j++) {
> + bo = to_vc4_bo(&exec->bo[j]->base);
> + ww_mutex_unlock(&bo->resv->lock);
> + }
> +
> + if (contended_lock != -1 && contended_lock >= i) {
> + bo = to_vc4_bo(&exec->bo[contended_lock]->base);
> +
> + ww_mutex_unlock(&bo->resv->lock);
> + }
> +
> + if (ret == -EDEADLK) {
> + contended_lock = i;
> + goto retry;
> + }
> +
> + ww_acquire_done(acquire_ctx);
> + return ret;
> + }
> + }
> +
> + ww_acquire_done(acquire_ctx);
> +
> + /* Reserve space for our shared (read-only) fence references,
> + * before we commit the CL to the hardware.
> + */
> + for (i = 0; i < exec->bo_count; i++) {
> + bo = to_vc4_bo(&exec->bo[i]->base);
> +
> + ret = reservation_object_reserve_shared(bo->resv);
> + if (ret) {
> + vc4_unlock_bo_reservations(dev, exec, acquire_ctx);
> + return ret;
> + }
> + }
> +
> + return 0;
> }
>
> /* Queues a struct vc4_exec_info for execution. If no job is
> @@ -484,19 +582,34 @@ vc4_update_bo_seqnos(struct vc4_exec_info *exec, uint64_t seqno)
> * then bump the end address. That's a change for a later date,
> * though.
> */
> -static void
> -vc4_queue_submit(struct drm_device *dev, struct vc4_exec_info *exec)
> +static int
> +vc4_queue_submit(struct drm_device *dev, struct vc4_exec_info *exec,
> + struct ww_acquire_ctx *acquire_ctx)
> {
> struct vc4_dev *vc4 = to_vc4_dev(dev);
> uint64_t seqno;
> unsigned long irqflags;
> + struct vc4_fence *fence;
> +
> + fence = kzalloc(sizeof(*fence), GFP_KERNEL);
> + if (!fence)
> + return -ENOMEM;
> + fence->dev = dev;
>
> spin_lock_irqsave(&vc4->job_lock, irqflags);
>
> seqno = ++vc4->emit_seqno;
> exec->seqno = seqno;
> +
> + dma_fence_init(&fence->base, &vc4_fence_ops, &vc4->job_lock,
> + vc4->dma_fence_context, exec->seqno);
> + fence->seqno = exec->seqno;
> + exec->fence = &fence->base;
> +
> vc4_update_bo_seqnos(exec, seqno);
>
> + vc4_unlock_bo_reservations(dev, exec, acquire_ctx);
> +
> list_add_tail(&exec->head, &vc4->bin_job_list);
>
> /* If no job was executing, kick ours off. Otherwise, it'll
> @@ -509,6 +622,8 @@ vc4_queue_submit(struct drm_device *dev, struct vc4_exec_info *exec)
> }
>
> spin_unlock_irqrestore(&vc4->job_lock, irqflags);
> +
> + return 0;
> }
>
> /**
> @@ -707,6 +822,12 @@ vc4_complete_exec(struct drm_device *dev, struct vc4_exec_info *exec)
> struct vc4_dev *vc4 = to_vc4_dev(dev);
> unsigned i;
>
> + /* If we got force-completed because of GPU reset rather than
> + * through our IRQ handler, signal the fence now.
> + */
> + if (exec->fence)
> + dma_fence_signal(exec->fence);
> +
> if (exec->bo) {
> for (i = 0; i < exec->bo_count; i++)
> drm_gem_object_unreference_unlocked(&exec->bo[i]->base);
> @@ -874,6 +995,7 @@ vc4_submit_cl_ioctl(struct drm_device *dev, void *data,
> struct vc4_dev *vc4 = to_vc4_dev(dev);
> struct drm_vc4_submit_cl *args = data;
> struct vc4_exec_info *exec;
> + struct ww_acquire_ctx acquire_ctx;
> int ret = 0;
>
> if ((args->flags & ~VC4_SUBMIT_CL_USE_CLEAR_COLOR) != 0) {
> @@ -916,12 +1038,18 @@ vc4_submit_cl_ioctl(struct drm_device *dev, void *data,
> if (ret)
> goto fail;
>
> + ret = vc4_lock_bo_reservations(dev, exec, &acquire_ctx);
> + if (ret)
> + goto fail;
> +
> /* Clear this out of the struct we'll be putting in the queue,
> * since it's part of our stack.
> */
> exec->args = NULL;
>
> - vc4_queue_submit(dev, exec);
> + ret = vc4_queue_submit(dev, exec, &acquire_ctx);
> + if (ret)
> + goto fail;
>
> /* Return the seqno for our job. */
> args->seqno = vc4->emit_seqno;
> @@ -939,6 +1067,8 @@ vc4_gem_init(struct drm_device *dev)
> {
> struct vc4_dev *vc4 = to_vc4_dev(dev);
>
> + vc4->dma_fence_context = dma_fence_context_alloc(1);
> +
> INIT_LIST_HEAD(&vc4->bin_job_list);
> INIT_LIST_HEAD(&vc4->render_job_list);
> INIT_LIST_HEAD(&vc4->job_done_list);
> diff --git a/drivers/gpu/drm/vc4/vc4_irq.c b/drivers/gpu/drm/vc4/vc4_irq.c
> index cdc6e6760705..1384af9fc987 100644
> --- a/drivers/gpu/drm/vc4/vc4_irq.c
> +++ b/drivers/gpu/drm/vc4/vc4_irq.c
> @@ -142,6 +142,10 @@ vc4_irq_finish_render_job(struct drm_device *dev)
>
> vc4->finished_seqno++;
> list_move_tail(&exec->head, &vc4->job_done_list);
> + if (exec->fence) {
> + dma_fence_signal_locked(exec->fence);
> + exec->fence = NULL;
> + }
> vc4_submit_next_render_job(dev);
>
> wake_up_all(&vc4->job_wait_queue);
> --
> 2.11.0
>
> _______________________________________________
> dri-devel mailing list
> dri-devel@xxxxxxxxxxxxxxxxxxxxx
> https://lists.freedesktop.org/mailman/listinfo/dri-devel

--
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch