[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20180118173814.GC9598@jade>
Date: Thu, 18 Jan 2018 15:38:14 -0200
From: Gustavo Padovan <gustavo@...ovan.org>
To: Hans Verkuil <hverkuil@...all.nl>
Cc: linux-media@...r.kernel.org,
Mauro Carvalho Chehab <mchehab@....samsung.com>,
Shuah Khan <shuahkh@....samsung.com>,
Pawel Osciak <pawel@...iak.com>,
Alexandre Courbot <acourbot@...omium.org>,
Sakari Ailus <sakari.ailus@....fi>,
Brian Starkey <brian.starkey@....com>,
Thierry Escande <thierry.escande@...labora.com>,
linux-kernel@...r.kernel.org,
Gustavo Padovan <gustavo.padovan@...labora.com>
Subject: Re: [PATCH v7 4/6] [media] vb2: add in-fence support to QBUF
Hi Hans,
2018-01-12 Hans Verkuil <hverkuil@...all.nl>:
> On 01/10/18 17:07, Gustavo Padovan wrote:
> > From: Gustavo Padovan <gustavo.padovan@...labora.com>
> >
> > Receive in-fence from userspace and add support for waiting on them
> > before queueing the buffer to the driver. Buffers can't be queued to the
> > driver before its fences signal. And a buffer can't be queue to the driver
> > out of the order they were queued from userspace. That means that even if
> > it fence signal it must wait all other buffers, ahead of it in the queue,
> > to signal first.
> >
> > If the fence for some buffer fails we do not queue it to the driver,
> > instead we mark it as error and wait until the previous buffer is done
> > to notify userspace of the error. We wait here to deliver the buffers back
> > to userspace in order.
> >
> > v8: - improve comments about fences with errors
> >
> > v7:
> > - get rid of the fence array stuff for ordering and just use
> > get_num_buffers_ready() (Hans)
> > - fix issue of queuing the buffer twice (Hans)
> > - avoid the dma_fence_wait() in core_qbuf() (Alex)
> > - merge preparation commit in
> >
> > v6:
> > - With fences always keep the order userspace queues the buffers.
> > - Protect in_fence manipulation with a lock (Brian Starkey)
> > - check if fences have the same context before adding a fence array
> > - Fix last_fence ref unbalance in __set_in_fence() (Brian Starkey)
> > - Clean up fence if __set_in_fence() fails (Brian Starkey)
> > - treat -EINVAL from dma_fence_add_callback() (Brian Starkey)
> >
> > v5: - use fence_array to keep buffers ordered in vb2 core when
> > needed (Brian Starkey)
> > - keep backward compat on the reserved2 field (Brian Starkey)
> > - protect fence callback removal with lock (Brian Starkey)
> >
> > v4:
> > - Add a comment about dma_fence_add_callback() not returning a
> > error (Hans)
> > - Call dma_fence_put(vb->in_fence) if fence signaled (Hans)
> > - select SYNC_FILE under config VIDEOBUF2_CORE (Hans)
> > - Move dma_fence_is_signaled() check to __enqueue_in_driver() (Hans)
> > - Remove list_for_each_entry() in __vb2_core_qbuf() (Hans)
> > - Remove if (vb->state != VB2_BUF_STATE_QUEUED) from
> > vb2_start_streaming() (Hans)
> > - set IN_FENCE flags on __fill_v4l2_buffer (Hans)
> > - Queue buffers to the driver as soon as they are ready (Hans)
> > - call fill_user_buffer() after queuing the buffer (Hans)
> > - add err: label to clean up fence
> > - add dma_fence_wait() before calling vb2_start_streaming()
> >
> > v3: - document fence parameter
> > - remove ternary if at vb2_qbuf() return (Mauro)
> > - do not change if conditions behaviour (Mauro)
> >
> > v2:
> > - fix vb2_queue_or_prepare_buf() ret check
> > - remove check for VB2_MEMORY_DMABUF only (Javier)
> > - check num of ready buffers to start streaming
> > - when queueing, start from the first ready buffer
> > - handle queue cancel
> >
> > Signed-off-by: Gustavo Padovan <gustavo.padovan@...labora.com>
> > ---
> > drivers/media/common/videobuf/videobuf2-core.c | 166 ++++++++++++++++++++++---
> > drivers/media/common/videobuf/videobuf2-v4l2.c | 29 ++++-
> > drivers/media/v4l2-core/Kconfig | 33 +++++
> > include/media/videobuf2-core.h | 14 ++-
> > 4 files changed, 221 insertions(+), 21 deletions(-)
> >
> > diff --git a/drivers/media/common/videobuf/videobuf2-core.c b/drivers/media/common/videobuf/videobuf2-core.c
> > index f7109f827f6e..777e3a2bc746 100644
> > --- a/drivers/media/common/videobuf/videobuf2-core.c
> > +++ b/drivers/media/common/videobuf/videobuf2-core.c
> > @@ -352,6 +352,7 @@ static int __vb2_queue_alloc(struct vb2_queue *q, enum vb2_memory memory,
> > vb->index = q->num_buffers + buffer;
> > vb->type = q->type;
> > vb->memory = memory;
> > + spin_lock_init(&vb->fence_cb_lock);
> > for (plane = 0; plane < num_planes; ++plane) {
> > vb->planes[plane].length = plane_sizes[plane];
> > vb->planes[plane].min_length = plane_sizes[plane];
> > @@ -936,7 +937,7 @@ void vb2_buffer_done(struct vb2_buffer *vb, enum vb2_buffer_state state)
> >
> > switch (state) {
> > case VB2_BUF_STATE_QUEUED:
> > - return;
> > + break;
> > case VB2_BUF_STATE_REQUEUEING:
> > if (q->start_streaming_called)
> > __enqueue_in_driver(vb);
> > @@ -946,6 +947,19 @@ void vb2_buffer_done(struct vb2_buffer *vb, enum vb2_buffer_state state)
> > wake_up(&q->done_wq);
> > break;
> > }
> > +
> > + /*
> > + * The check below verifies if there is a buffer in queue with an
>
> s/in/in the/
>
> > + * error state. They are added to queue in the error state when
> > + * their in-fence fails to signal.
> > + * To not mess with buffer ordering we wait until the previous buffer
> > + * is done to mark the buffer in the error state as done and notify
> > + * userspace. So everytime a buffer is done we check the next one for
>
> s/everytime/every time/
>
> > + * VB2_BUF_STATE_ERROR.
> > + */
> > + vb = list_next_entry(vb, queued_entry);
> > + if (vb && vb->state == VB2_BUF_STATE_ERROR)
> > + vb2_buffer_done(vb, vb->state);
>
> Hmm. I'm not sure this is correct. Please test this and make sure you have
> enabled CONFIG_VIDEO_ADV_DEBUG. This enables additional instrumentation that
> will warn if the internal vb2 state becomes unbalanced.
>
> An additional problem I have with this recursive call is what happens when
> a whole bunch of in fences signal an error, so you have e.g. 10 consecutive
> buffers with state ERROR. You'd recurse 10 times in that case.
>
> I think you are better off handling this corner case explicitly, e.g. something
> like this:
>
> for (;;) {
> vb = list_next_entry(vb, queued_entry);
> if (!vb || vb->state != VB2_BUF_STATE_ERROR)
> break;
>
> dprintk(4, "done processing on buffer %d, state: %d\n",
> vb->index, vb->state);
>
> if (state == VB2_BUF_STATE_QUEUED) {
> vb->state = state;
> } else {
> spin_lock_irqsave(&q->done_lock, flags);
> /* Add the buffer to the done buffers list */
> list_add_tail(&vb->done_entry, &q->done_list);
> spin_unlock_irqrestore(&q->done_lock, flags);
>
> /* Inform any processes that may be waiting for buffers */
> wake_up(&q->done_wq);
> }
>
> trace_vb2_buf_done(q, vb);
> }
>
> I *think* this ensures all counters remain balanced, but this really needs
> to be tested. I haven't compiled this code, so I hope I got it right.
I'll play with this and make sure I get it right.
>
> > }
> > EXPORT_SYMBOL_GPL(vb2_buffer_done);
> >
> > @@ -1230,6 +1244,9 @@ static void __enqueue_in_driver(struct vb2_buffer *vb)
> > {
> > struct vb2_queue *q = vb->vb2_queue;
> >
> > + if (vb->in_fence && !dma_fence_is_signaled(vb->in_fence))
> > + return;
> > +
> > vb->state = VB2_BUF_STATE_ACTIVE;
> > atomic_inc(&q->owned_by_drv_count);
> >
> > @@ -1281,6 +1298,24 @@ static int __buf_prepare(struct vb2_buffer *vb, const void *pb)
> > return 0;
> > }
> >
> > +static int __get_num_ready_buffers(struct vb2_queue *q)
> > +{
> > + struct vb2_buffer *vb;
> > + int ready_count = 0;
> > + unsigned long flags;
> > +
> > + /* count num of buffers ready in front of the queued_list */
> > + list_for_each_entry(vb, &q->queued_list, queued_entry) {
> > + spin_lock_irqsave(&vb->fence_cb_lock, flags);
> > + if (vb->in_fence && !dma_fence_is_signaled(vb->in_fence))
> > + break;
> > + ready_count++;
> > + spin_unlock_irqrestore(&vb->fence_cb_lock, flags);
> > + }
> > +
> > + return ready_count;
> > +}
> > +
> > int vb2_core_prepare_buf(struct vb2_queue *q, unsigned int index, void *pb)
> > {
> > struct vb2_buffer *vb;
> > @@ -1369,9 +1404,43 @@ static int vb2_start_streaming(struct vb2_queue *q)
> > return ret;
> > }
> >
> > -int vb2_core_qbuf(struct vb2_queue *q, unsigned int index, void *pb)
> > +static void vb2_qbuf_fence_cb(struct dma_fence *f, struct dma_fence_cb *cb)
> > +{
> > + struct vb2_buffer *vb = container_of(cb, struct vb2_buffer, fence_cb);
> > + struct vb2_queue *q = vb->vb2_queue;
> > + unsigned long flags;
> > +
> > + spin_lock_irqsave(&vb->fence_cb_lock, flags);
> > + /*
> > + * If the fence signal with an error we mark the buffer as such
>
> s/signal/signals/
>
> > + * and avoid using it by setting it to VB2_BUF_STATE_ERROR and
> > + * not queueing it to the driver. However we can't notify the error
> > + * to userspace right now because, at the time this callback run, QBUF
> > + * returned already.
> > + * So we delay that to DQBUF time. See comments in vb2_buffer_done()
> > + * as well.
> > + */
> > + if (vb->in_fence->error)
> > + vb->state = VB2_BUF_STATE_ERROR;
> > +
> > + dma_fence_put(vb->in_fence);
> > + vb->in_fence = NULL;
> > +
> > + if (vb->state == VB2_BUF_STATE_ERROR) {
> > + spin_unlock_irqrestore(&vb->fence_cb_lock, flags);
> > + return;
> > + }
> > +
> > + if (q->start_streaming_called)
> > + __enqueue_in_driver(vb);
> > + spin_unlock_irqrestore(&vb->fence_cb_lock, flags);
> > +}
> > +
> > +int vb2_core_qbuf(struct vb2_queue *q, unsigned int index, void *pb,
> > + struct dma_fence *fence)
> > {
> > struct vb2_buffer *vb;
> > + unsigned long flags;
> > int ret;
> >
> > vb = q->bufs[index];
> > @@ -1380,16 +1449,18 @@ int vb2_core_qbuf(struct vb2_queue *q, unsigned int index, void *pb)
> > case VB2_BUF_STATE_DEQUEUED:
> > ret = __buf_prepare(vb, pb);
> > if (ret)
> > - return ret;
> > + goto err;
> > break;
> > case VB2_BUF_STATE_PREPARED:
> > break;
> > case VB2_BUF_STATE_PREPARING:
> > dprintk(1, "buffer still being prepared\n");
> > - return -EINVAL;
> > + ret = -EINVAL;
> > + goto err;
> > default:
> > dprintk(1, "invalid buffer state %d\n", vb->state);
> > - return -EINVAL;
> > + ret = -EINVAL;
> > + goto err;
> > }
> >
> > /*
> > @@ -1400,6 +1471,7 @@ int vb2_core_qbuf(struct vb2_queue *q, unsigned int index, void *pb)
> > q->queued_count++;
> > q->waiting_for_buffers = false;
> > vb->state = VB2_BUF_STATE_QUEUED;
> > + vb->in_fence = fence;
> >
> > if (pb)
> > call_void_bufop(q, copy_timestamp, vb, pb);
> > @@ -1407,15 +1479,42 @@ int vb2_core_qbuf(struct vb2_queue *q, unsigned int index, void *pb)
> > trace_vb2_qbuf(q, vb);
> >
> > /*
> > - * If already streaming, give the buffer to driver for processing.
> > - * If not, the buffer will be given to driver on next streamon.
> > + * For explicit synchronization: If the fence didn't signal
> > + * yet we setup a callback to queue the buffer once the fence
> > + * signals, and then, return successfully. But if the fence
> > + * already signaled we lose the reference we held and queue the
> > + * buffer to the driver.
> > */
> > - if (q->start_streaming_called)
> > - __enqueue_in_driver(vb);
> > + spin_lock_irqsave(&vb->fence_cb_lock, flags);
> > + if (vb->in_fence) {
> > + ret = dma_fence_add_callback(vb->in_fence, &vb->fence_cb,
> > + vb2_qbuf_fence_cb);
> > + if (ret == -EINVAL) {
> > + spin_unlock_irqrestore(&vb->fence_cb_lock, flags);
> > + goto err;
> > + } else if (!ret) {
> > + goto fill;
> > + }
> >
> > - /* Fill buffer information for the userspace */
> > - if (pb)
> > - call_void_bufop(q, fill_user_buffer, vb, pb);
> > + dma_fence_put(vb->in_fence);
> > + vb->in_fence = NULL;
> > + }
> > +
> > +fill:
> > + /*
> > + * If already streaming and there is no fence to wait on
> > + * give the buffer to driver for processing.
> > + */
> > + if (q->start_streaming_called) {
> > + struct vb2_buffer *b;
>
> Add empty line.
>
> > + list_for_each_entry(b, &q->queued_list, queued_entry) {
> > + if (b->state != VB2_BUF_STATE_QUEUED)
> > + continue;
> > + if (b->in_fence)
> > + break;
> > + __enqueue_in_driver(b);
> > + }
> > + }
> >
> > /*
> > * If streamon has been called, and we haven't yet called
> > @@ -1424,14 +1523,33 @@ int vb2_core_qbuf(struct vb2_queue *q, unsigned int index, void *pb)
> > * then we can finally call start_streaming().
> > */
> > if (q->streaming && !q->start_streaming_called &&
> > - q->queued_count >= q->min_buffers_needed) {
> > + __get_num_ready_buffers(q) >= q->min_buffers_needed) {
> > ret = vb2_start_streaming(q);
> > if (ret)
> > - return ret;
> > + goto err;
>
> You're missing a spin_unlock_irqrestore() call in this error path.
>
> > }
> >
> > + spin_unlock_irqrestore(&vb->fence_cb_lock, flags);
> > +
> > + /* Fill buffer information for the userspace */
> > + if (pb)
> > + call_void_bufop(q, fill_user_buffer, vb, pb);
> > +
> > dprintk(2, "qbuf of buffer %d succeeded\n", vb->index);
> > return 0;
> > +
> > +err:
> > + /* Fill buffer information for the userspace */
> > + if (pb)
> > + call_void_bufop(q, fill_user_buffer, vb, pb);
> > +
> > + if (vb->in_fence) {
> > + dma_fence_put(vb->in_fence);
> > + vb->in_fence = NULL;
> > + }
> > +
> > + return ret;
> > +
> > }
> > EXPORT_SYMBOL_GPL(vb2_core_qbuf);
> >
> > @@ -1642,6 +1760,8 @@ EXPORT_SYMBOL_GPL(vb2_core_dqbuf);
> > static void __vb2_queue_cancel(struct vb2_queue *q)
> > {
> > unsigned int i;
> > + struct vb2_buffer *vb;
> > + unsigned long flags;
> >
> > /*
> > * Tell driver to stop all transactions and release all queued
> > @@ -1672,6 +1792,16 @@ static void __vb2_queue_cancel(struct vb2_queue *q)
> > q->queued_count = 0;
> > q->error = 0;
> >
> > + list_for_each_entry(vb, &q->queued_list, queued_entry) {
> > + spin_lock_irqsave(&vb->fence_cb_lock, flags);
> > + if (vb->in_fence) {
> > + dma_fence_remove_callback(vb->in_fence, &vb->fence_cb);
> > + dma_fence_put(vb->in_fence);
> > + vb->in_fence = NULL;
> > + }
> > + spin_unlock_irqrestore(&vb->fence_cb_lock, flags);
> > + }
> > +
> > /*
> > * Remove all buffers from videobuf's list...
> > */
> > @@ -1733,7 +1863,7 @@ int vb2_core_streamon(struct vb2_queue *q, unsigned int type)
> > * Tell driver to start streaming provided sufficient buffers
> > * are available.
> > */
> > - if (q->queued_count >= q->min_buffers_needed) {
> > + if (__get_num_ready_buffers(q) >= q->min_buffers_needed) {
> > ret = v4l_vb2q_enable_media_source(q);
> > if (ret)
> > return ret;
> > @@ -2255,7 +2385,7 @@ static int __vb2_init_fileio(struct vb2_queue *q, int read)
> > * Queue all buffers.
> > */
> > for (i = 0; i < q->num_buffers; i++) {
> > - ret = vb2_core_qbuf(q, i, NULL);
> > + ret = vb2_core_qbuf(q, i, NULL, NULL);
> > if (ret)
> > goto err_reqbufs;
> > fileio->bufs[i].queued = 1;
> > @@ -2434,7 +2564,7 @@ static size_t __vb2_perform_fileio(struct vb2_queue *q, char __user *data, size_
> >
> > if (copy_timestamp)
> > b->timestamp = ktime_get_ns();
> > - ret = vb2_core_qbuf(q, index, NULL);
> > + ret = vb2_core_qbuf(q, index, NULL, NULL);
> > dprintk(5, "vb2_dbuf result: %d\n", ret);
> > if (ret)
> > return ret;
> > @@ -2537,7 +2667,7 @@ static int vb2_thread(void *data)
> > if (copy_timestamp)
> > vb->timestamp = ktime_get_ns();
> > if (!threadio->stop)
> > - ret = vb2_core_qbuf(q, vb->index, NULL);
> > + ret = vb2_core_qbuf(q, vb->index, NULL, NULL);
> > call_void_qop(q, wait_prepare, q);
> > if (ret || threadio->stop)
> > break;
> > diff --git a/drivers/media/common/videobuf/videobuf2-v4l2.c b/drivers/media/common/videobuf/videobuf2-v4l2.c
> > index d838524a459e..0a41e3bb7733 100644
> > --- a/drivers/media/common/videobuf/videobuf2-v4l2.c
> > +++ b/drivers/media/common/videobuf/videobuf2-v4l2.c
> > @@ -23,6 +23,7 @@
> > #include <linux/sched.h>
> > #include <linux/freezer.h>
> > #include <linux/kthread.h>
> > +#include <linux/sync_file.h>
> >
> > #include <media/v4l2-dev.h>
> > #include <media/v4l2-fh.h>
> > @@ -178,6 +179,12 @@ static int vb2_queue_or_prepare_buf(struct vb2_queue *q, struct v4l2_buffer *b,
> > return -EINVAL;
> > }
> >
> > + if ((b->fence_fd != 0 && b->fence_fd != -1) &&
>
> Wouldn't 'b->fence_fd > 0' be more sensible?
This is the case where the in fence flag is not set, I think we also
want to cover < -1 values userspace may set.
Gustavo
Powered by blists - more mailing lists