// Copyright 2023 The Chromium Authors
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.
#include "media/gpu/v4l2/v4l2_queue.h"
#include <errno.h>
#include <fcntl.h>
#include <libdrm/drm_fourcc.h>
#include <linux/media.h>
#include <poll.h>
#include <sys/ioctl.h>
#include <sys/mman.h>
#include "base/containers/contains.h"
#include "base/not_fatal_until.h"
#include "base/posix/eintr_wrapper.h"
#include "base/trace_event/trace_event.h"
#include "media/gpu/chromeos/native_pixmap_frame_resource.h"
#include "media/gpu/chromeos/platform_video_frame_utils.h"
#include "media/gpu/macros.h"
namespace media {
namespace {
// TODO(jkardatzke): Remove this when it is in linux/videodev2.h.
#define V4L2_MEMORY_FLAG_SECURE 0x2
// Maximum number of requests that can be created.
constexpr size_t kMaxNumRequests = 32;
gfx::Rect V4L2RectToGfxRect(const v4l2_rect& rect) {
return gfx::Rect(rect.left, rect.top, rect.width, rect.height);
}
struct v4l2_format BuildV4L2Format(const enum v4l2_buf_type type,
uint32_t fourcc,
const gfx::Size& size,
size_t buffer_size) {
struct v4l2_format format;
memset(&format, 0, sizeof(format));
format.type = type;
format.fmt.pix_mp.pixelformat = fourcc;
format.fmt.pix_mp.width = size.width();
format.fmt.pix_mp.height = size.height();
format.fmt.pix_mp.num_planes = GetNumPlanesOfV4L2PixFmt(fourcc);
format.fmt.pix_mp.plane_fmt[0].sizeimage = buffer_size;
return format;
}
const char* V4L2BufferTypeToString(const enum v4l2_buf_type buf_type) {
switch (buf_type) {
case V4L2_BUF_TYPE_VIDEO_OUTPUT:
return "OUTPUT";
case V4L2_BUF_TYPE_VIDEO_CAPTURE:
return "CAPTURE";
case V4L2_BUF_TYPE_VIDEO_OUTPUT_MPLANE:
return "OUTPUT_MPLANE";
case V4L2_BUF_TYPE_VIDEO_CAPTURE_MPLANE:
return "CAPTURE_MPLANE";
default:
return "UNKNOWN";
}
}
int64_t V4L2BufferTimestampInMilliseconds(
const struct v4l2_buffer* v4l2_buffer) {
struct timespec ts;
TIMEVAL_TO_TIMESPEC(&v4l2_buffer->timestamp, &ts);
return base::TimeDelta::FromTimeSpec(ts).InMilliseconds();
}
// For decoding and encoding data to be processed is enqueued in the
// V4L2_BUF_TYPE_VIDEO_OUTPUT_MPLANE queue. Once that data has been either
// decompressed or compressed, the finished buffer is dequeued from the
// V4L2_BUF_TYPE_VIDEO_CAPTURE_MPLANE queue. This occurs asynchronously so
// there is no way to measure how long the hardware took to process the data.
// We can use the length of time that a buffer is enqueued as a proxy for
// how busy the hardware is.
void V4L2ProcessingTrace(const struct v4l2_buffer* v4l2_buffer, bool start) {
constexpr char kTracingCategory[] = "media,gpu";
constexpr char kQueueBuffer[] = "V4L2 Queue Buffer";
constexpr char kDequeueBuffer[] = "V4L2 Dequeue Buffer";
constexpr char kVideoDecoding[] = "V4L2 Video Decoding";
bool tracing_enabled = false;
TRACE_EVENT_CATEGORY_GROUP_ENABLED(kTracingCategory, &tracing_enabled);
if (!tracing_enabled) {
return;
}
const char* name = start ? kQueueBuffer : kDequeueBuffer;
TRACE_EVENT_INSTANT1(kTracingCategory, name, TRACE_EVENT_SCOPE_THREAD, "type",
v4l2_buffer->type);
// TODO(mcasas): Consider using TimeValToTimeDelta().
const int64_t timestamp = V4L2BufferTimestampInMilliseconds(v4l2_buffer);
if (timestamp <= 0) {
return;
}
if (start && v4l2_buffer->type == V4L2_BUF_TYPE_VIDEO_OUTPUT_MPLANE) {
TRACE_EVENT_NESTABLE_ASYNC_BEGIN1(kTracingCategory, kVideoDecoding,
TRACE_ID_LOCAL(timestamp), "timestamp",
timestamp);
} else if (!start &&
v4l2_buffer->type == V4L2_BUF_TYPE_VIDEO_CAPTURE_MPLANE) {
TRACE_EVENT_NESTABLE_ASYNC_END1(kTracingCategory, kVideoDecoding,
TRACE_ID_LOCAL(timestamp), "timestamp",
timestamp);
}
}
// Returns a vector of dmabuf file descriptors, exported for V4L2 buffer with
// |index|, assuming the buffer contains |num_planes| V4L2 planes and is of
// |buf_type|. Returns an empty vector on failure. The caller is responsible for
// closing the file descriptors after use.
std::vector<base::ScopedFD> GetDmabufsForV4L2Buffer(
const IoctlAsCallback& ioctl_cb,
int index,
size_t num_planes,
enum v4l2_buf_type buf_type) {
DVLOGF(3);
DCHECK(V4L2_TYPE_IS_MULTIPLANAR(buf_type));
std::vector<base::ScopedFD> dmabuf_fds;
for (size_t i = 0; i < num_planes; ++i) {
struct v4l2_exportbuffer expbuf;
memset(&expbuf, 0, sizeof(expbuf));
expbuf.type = buf_type;
expbuf.index = index;
expbuf.plane = i;
expbuf.flags = O_CLOEXEC;
if (ioctl_cb.Run(VIDIOC_EXPBUF, &expbuf) != 0) {
RecordVidiocIoctlErrorUMA(VidiocIoctlRequests::kVidiocExpbuf);
dmabuf_fds.clear();
break;
}
dmabuf_fds.push_back(base::ScopedFD(expbuf.fd));
}
return dmabuf_fds;
}
} // namespace
V4L2ExtCtrl::V4L2ExtCtrl(uint32_t id) {
memset(&ctrl, 0, sizeof(ctrl));
ctrl.id = id;
}
V4L2ExtCtrl::V4L2ExtCtrl(uint32_t id, int32_t val) : V4L2ExtCtrl(id) {
ctrl.value = val;
}
// Class used to store the state of a buffer that should persist between
// reference creations. This includes:
// * Result of initial VIDIOC_QUERYBUF ioctl,
// * Plane mappings.
//
// Also provides helper functions.
class V4L2Buffer {
public:
static std::unique_ptr<V4L2Buffer> Create(
const IoctlAsCallback& ioctl_cb,
const MmapAsCallback& mmap_cb,
enum v4l2_buf_type type,
enum v4l2_memory memory,
const struct v4l2_format& format,
size_t buffer_id);
V4L2Buffer(const V4L2Buffer&) = delete;
V4L2Buffer& operator=(const V4L2Buffer&) = delete;
~V4L2Buffer();
void* GetPlaneMapping(const size_t plane);
size_t GetMemoryUsage() const;
const struct v4l2_buffer& v4l2_buffer() const { return v4l2_buffer_; }
const scoped_refptr<FrameResource>& GetFrameResource();
private:
V4L2Buffer(const IoctlAsCallback& ioctl_cb,
const MmapAsCallback& mmap_cb,
enum v4l2_buf_type type,
enum v4l2_memory memory,
const struct v4l2_format& format,
size_t buffer_id);
bool Query();
scoped_refptr<FrameResource> CreateFrame();
const IoctlAsCallback ioctl_cb_;
const MmapAsCallback mmap_cb_;
std::vector<void*> plane_mappings_;
// V4L2 data as queried by QUERYBUF.
struct v4l2_buffer v4l2_buffer_;
// WARNING: do not change this to a vector or something smaller than
// VIDEO_MAX_PLANES (the maximum number of planes V4L2 supports). The
// element overhead is small and may avoid memory corruption bugs.
struct v4l2_plane v4l2_planes_[VIDEO_MAX_PLANES];
struct v4l2_format format_;
scoped_refptr<FrameResource> frame_;
base::WeakPtrFactory<V4L2Buffer> weak_factory_{this};
};
std::unique_ptr<V4L2Buffer> V4L2Buffer::Create(
const IoctlAsCallback& ioctl_cb,
const MmapAsCallback& mmap_cb,
enum v4l2_buf_type type,
enum v4l2_memory memory,
const struct v4l2_format& format,
size_t buffer_id) {
// Not using std::make_unique because constructor is private.
std::unique_ptr<V4L2Buffer> buffer(new V4L2Buffer(std::move(ioctl_cb),
std::move(mmap_cb), type,
memory, format, buffer_id));
if (!buffer->Query()) {
return nullptr;
}
return buffer;
}
V4L2Buffer::V4L2Buffer(const IoctlAsCallback& ioctl_cb,
const MmapAsCallback& mmap_cb,
enum v4l2_buf_type type,
enum v4l2_memory memory,
const struct v4l2_format& format,
size_t buffer_id)
: ioctl_cb_(ioctl_cb), mmap_cb_(mmap_cb), format_(format) {
DCHECK(V4L2_TYPE_IS_MULTIPLANAR(type));
DCHECK_LE(format.fmt.pix_mp.num_planes, std::size(v4l2_planes_));
memset(&v4l2_buffer_, 0, sizeof(v4l2_buffer_));
memset(v4l2_planes_, 0, sizeof(v4l2_planes_));
v4l2_buffer_.m.planes = v4l2_planes_;
// Just in case we got more planes than we want.
v4l2_buffer_.length =
std::min(static_cast<size_t>(format.fmt.pix_mp.num_planes),
std::size(v4l2_planes_));
v4l2_buffer_.index = buffer_id;
v4l2_buffer_.type = type;
v4l2_buffer_.memory = memory;
plane_mappings_.resize(v4l2_buffer_.length);
}
V4L2Buffer::~V4L2Buffer() {
if (v4l2_buffer_.memory == V4L2_MEMORY_MMAP) {
for (size_t i = 0; i < plane_mappings_.size(); i++) {
if (plane_mappings_[i] != nullptr) {
munmap(plane_mappings_[i], v4l2_buffer_.m.planes[i].length);
}
}
}
}
bool V4L2Buffer::Query() {
int ret = ioctl_cb_.Run(VIDIOC_QUERYBUF, &v4l2_buffer_);
if (ret) {
RecordVidiocIoctlErrorUMA(VidiocIoctlRequests::kVidiocQuerybuf);
VPLOGF(1) << "VIDIOC_QUERYBUF failed: ";
return false;
}
DCHECK(plane_mappings_.size() == v4l2_buffer_.length);
return true;
}
void* V4L2Buffer::GetPlaneMapping(const size_t plane) {
if (plane >= plane_mappings_.size()) {
VLOGF(1) << "Invalid plane " << plane << " requested.";
return nullptr;
}
void* p = plane_mappings_[plane];
if (p) {
return p;
}
// Do this check here to avoid repeating it after a buffer has been
// successfully mapped (we know we are of MMAP type by then).
if (v4l2_buffer_.memory != V4L2_MEMORY_MMAP) {
VLOGF(1) << "Cannot create mapping on non-MMAP buffer";
return nullptr;
}
p = mmap_cb_.Run(nullptr, v4l2_buffer_.m.planes[plane].length,
PROT_READ | PROT_WRITE, MAP_SHARED,
v4l2_buffer_.m.planes[plane].m.mem_offset);
if (p == MAP_FAILED) {
VPLOGF(1) << "mmap() failed: ";
return nullptr;
}
plane_mappings_[plane] = p;
return p;
}
size_t V4L2Buffer::GetMemoryUsage() const {
size_t usage = 0;
for (size_t i = 0; i < v4l2_buffer_.length; i++) {
usage += v4l2_buffer_.m.planes[i].length;
}
return usage;
}
scoped_refptr<FrameResource> V4L2Buffer::CreateFrame() {
auto layout = V4L2FormatToVideoFrameLayout(format_);
if (!layout) {
VLOGF(1) << "Cannot create frame layout for V4L2 buffers";
return nullptr;
}
std::vector<base::ScopedFD> dmabuf_fds = GetDmabufsForV4L2Buffer(
ioctl_cb_, v4l2_buffer_.index, v4l2_buffer_.length,
static_cast<enum v4l2_buf_type>(v4l2_buffer_.type));
if (dmabuf_fds.empty()) {
VLOGF(1) << "Failed to get DMABUFs of V4L2 buffer";
return nullptr;
}
// DMA buffer fds should not be invalid
for (const auto& dmabuf_fd : dmabuf_fds) {
if (!dmabuf_fd.is_valid()) {
DLOG(ERROR) << "Fail to get DMABUFs of V4L2 buffer - invalid fd";
return nullptr;
}
}
// Duplicate the fd of the last v4l2 plane until the number of fds are the
// same as the number of color planes.
while (dmabuf_fds.size() < layout->planes().size()) {
int duped_fd = HANDLE_EINTR(dup(dmabuf_fds.back().get()));
if (duped_fd == -1) {
DLOG(ERROR) << "Failed duplicating dmabuf fd";
return nullptr;
}
dmabuf_fds.emplace_back(duped_fd);
}
gfx::Size size(format_.fmt.pix_mp.width, format_.fmt.pix_mp.height);
return NativePixmapFrameResource::Create(
*layout, gfx::Rect(size), size, std::move(dmabuf_fds), base::TimeDelta());
}
const scoped_refptr<FrameResource>& V4L2Buffer::GetFrameResource() {
// We can create the FrameResource only when using MMAP buffers.
if (v4l2_buffer_.memory != V4L2_MEMORY_MMAP) {
VLOGF(1) << "Cannot create video frame from non-MMAP buffer";
// Allow NOTREACHED() on invalid argument because this is an internal
// method.
NOTREACHED_IN_MIGRATION();
}
// Create the video frame instance if requiring it for the first time.
if (!frame_) {
frame_ = CreateFrame();
}
return frame_;
}
// A thread-safe pool of buffer indexes, allowing buffers to be obtained and
// returned from different threads. All the methods of this class are
// thread-safe. Users should keep a scoped_refptr to instances of this class
// in order to ensure the list remains alive as long as they need it.
class V4L2BuffersList : public base::RefCountedThreadSafe<V4L2BuffersList> {
public:
V4L2BuffersList() = default;
V4L2BuffersList(const V4L2BuffersList&) = delete;
V4L2BuffersList& operator=(const V4L2BuffersList&) = delete;
// Return a buffer to this list. Also can be called to set the initial pool
// of buffers.
// Note that it is illegal to return the same buffer twice.
void ReturnBuffer(size_t buffer_id);
// Get any of the buffers in the list. There is no order guarantee whatsoever.
std::optional<size_t> GetFreeBuffer();
// Get the buffer with specified index.
std::optional<size_t> GetFreeBuffer(size_t requested_buffer_id);
// Number of buffers currently in this list.
size_t size() const;
private:
friend class base::RefCountedThreadSafe<V4L2BuffersList>;
~V4L2BuffersList() = default;
mutable base::Lock lock_;
std::set<size_t> free_buffers_ GUARDED_BY(lock_);
};
void V4L2BuffersList::ReturnBuffer(size_t buffer_id) {
base::AutoLock auto_lock(lock_);
auto inserted = free_buffers_.emplace(buffer_id);
DCHECK(inserted.second);
}
std::optional<size_t> V4L2BuffersList::GetFreeBuffer() {
base::AutoLock auto_lock(lock_);
auto iter = free_buffers_.begin();
if (iter == free_buffers_.end()) {
DVLOGF(4) << "No free buffer available!";
return std::nullopt;
}
size_t buffer_id = *iter;
free_buffers_.erase(iter);
return buffer_id;
}
std::optional<size_t> V4L2BuffersList::GetFreeBuffer(
size_t requested_buffer_id) {
base::AutoLock auto_lock(lock_);
return (free_buffers_.erase(requested_buffer_id) > 0)
? std::make_optional(requested_buffer_id)
: std::nullopt;
}
size_t V4L2BuffersList::size() const {
base::AutoLock auto_lock(lock_);
return free_buffers_.size();
}
// Module-private class that let users query/write V4L2 buffer information.
// It also makes some private V4L2Queue methods available to this module only.
class V4L2BufferRefBase {
public:
V4L2BufferRefBase(const struct v4l2_buffer& v4l2_buffer,
base::WeakPtr<V4L2Queue> queue);
V4L2BufferRefBase(const V4L2BufferRefBase&) = delete;
V4L2BufferRefBase& operator=(const V4L2BufferRefBase&) = delete;
~V4L2BufferRefBase();
bool QueueBuffer(scoped_refptr<FrameResource> frame);
void* GetPlaneMapping(const size_t plane);
const scoped_refptr<FrameResource>& GetFrameResource();
// Checks that the number of passed FDs is adequate for the current format
// and buffer configuration. Only useful for DMABUF buffers.
bool CheckNumFDsForFormat(const size_t num_fds) const;
// Data from the buffer, that users can query and/or write.
struct v4l2_buffer v4l2_buffer_;
// WARNING: do not change this to a vector or something smaller than
// VIDEO_MAX_PLANES (the maximum number of planes V4L2 supports). The
// element overhead is small and may avoid memory corruption bugs.
struct v4l2_plane v4l2_planes_[VIDEO_MAX_PLANES];
private:
size_t BufferId() const { return v4l2_buffer_.index; }
friend class V4L2WritableBufferRef;
// A weak pointer to the queue this buffer belongs to. Will remain valid as
// long as the underlying V4L2 buffer is valid too.
// This can only be accessed from the sequence protected by sequence_checker_.
// Thread-safe methods (like ~V4L2BufferRefBase) must *never* access this.
base::WeakPtr<V4L2Queue> queue_;
// Where to return this buffer if it goes out of scope without being queued.
scoped_refptr<V4L2BuffersList> return_to_;
bool queued = false;
SEQUENCE_CHECKER(sequence_checker_);
};
V4L2BufferRefBase::V4L2BufferRefBase(const struct v4l2_buffer& v4l2_buffer,
base::WeakPtr<V4L2Queue> queue)
: queue_(std::move(queue)), return_to_(queue_->free_buffers_) {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK(V4L2_TYPE_IS_MULTIPLANAR(v4l2_buffer.type));
DCHECK_LE(v4l2_buffer.length, std::size(v4l2_planes_));
DCHECK(return_to_);
memcpy(&v4l2_buffer_, &v4l2_buffer, sizeof(v4l2_buffer_));
memcpy(v4l2_planes_, v4l2_buffer.m.planes,
sizeof(struct v4l2_plane) * v4l2_buffer.length);
v4l2_buffer_.m.planes = v4l2_planes_;
}
V4L2BufferRefBase::~V4L2BufferRefBase() {
// We are the last reference and are only accessing the thread-safe
// return_to_, so we are safe to call from any sequence.
// If we have been queued, then the queue is our owner so we don't need to
// return to the free buffers list.
if (!queued) {
return_to_->ReturnBuffer(BufferId());
}
}
bool V4L2BufferRefBase::QueueBuffer(scoped_refptr<FrameResource> frame) {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
if (!queue_) {
return false;
}
queued = queue_->QueueBuffer(&v4l2_buffer_, std::move(frame));
return queued;
}
void* V4L2BufferRefBase::GetPlaneMapping(const size_t plane) {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
if (!queue_) {
return nullptr;
}
return queue_->buffers_[BufferId()]->GetPlaneMapping(plane);
}
const scoped_refptr<FrameResource>& V4L2BufferRefBase::GetFrameResource() {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
// Used so we can return a const scoped_refptr& in all cases.
static const scoped_refptr<FrameResource> null_frame_resource;
if (!queue_) {
return null_frame_resource;
}
DCHECK_LE(BufferId(), queue_->buffers_.size());
return queue_->buffers_[BufferId()]->GetFrameResource();
}
bool V4L2BufferRefBase::CheckNumFDsForFormat(const size_t num_fds) const {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
if (!queue_) {
return false;
}
// We have not used SetFormat(), assume this is ok.
// Hopefully we standardize SetFormat() in the future.
if (!queue_->current_format_) {
return true;
}
const size_t required_fds = queue_->current_format_->fmt.pix_mp.num_planes;
// Sanity check.
DCHECK_EQ(v4l2_buffer_.length, required_fds);
if (num_fds < required_fds) {
VLOGF(1) << "Insufficient number of FDs given for the current format. "
<< num_fds << " provided, " << required_fds << " required.";
return false;
}
const auto* planes = v4l2_buffer_.m.planes;
for (size_t i = v4l2_buffer_.length - 1; i >= num_fds; --i) {
// Assume that an fd is a duplicate of a previous plane's fd if offset != 0.
// Otherwise, if offset == 0, return error as it is likely pointing to
// a new plane.
if (planes[i].data_offset == 0) {
VLOGF(1) << "Additional dmabuf fds point to a new buffer.";
return false;
}
}
return true;
}
V4L2WritableBufferRef::V4L2WritableBufferRef(
const struct v4l2_buffer& v4l2_buffer,
base::WeakPtr<V4L2Queue> queue)
: buffer_data_(
std::make_unique<V4L2BufferRefBase>(v4l2_buffer, std::move(queue))) {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
}
V4L2WritableBufferRef::V4L2WritableBufferRef(V4L2WritableBufferRef&& other)
: buffer_data_(std::move(other.buffer_data_)) {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK_CALLED_ON_VALID_SEQUENCE(other.sequence_checker_);
}
V4L2WritableBufferRef::~V4L2WritableBufferRef() {
// Only valid references should be sequence-checked
if (buffer_data_) {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
}
}
V4L2WritableBufferRef& V4L2WritableBufferRef::operator=(
V4L2WritableBufferRef&& other) {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK_CALLED_ON_VALID_SEQUENCE(other.sequence_checker_);
if (this == &other) {
return *this;
}
buffer_data_ = std::move(other.buffer_data_);
return *this;
}
scoped_refptr<FrameResource> V4L2WritableBufferRef::GetFrameResource() {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK(buffer_data_);
return buffer_data_->GetFrameResource();
}
enum v4l2_memory V4L2WritableBufferRef::Memory() const {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK(buffer_data_);
return static_cast<enum v4l2_memory>(buffer_data_->v4l2_buffer_.memory);
}
bool V4L2WritableBufferRef::DoQueue(V4L2RequestRef* request_ref,
scoped_refptr<FrameResource> frame) && {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK(buffer_data_);
if (request_ref && buffer_data_->queue_->SupportsRequests() &&
!request_ref->ApplyQueueBuffer(&(buffer_data_->v4l2_buffer_))) {
return false;
}
bool queued = buffer_data_->QueueBuffer(std::move(frame));
// Clear our own reference.
buffer_data_.reset();
return queued;
}
bool V4L2WritableBufferRef::QueueMMap(V4L2RequestRef* request_ref) && {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK(buffer_data_);
// Move ourselves so our data gets freed no matter when we return
V4L2WritableBufferRef self(std::move(*this));
if (self.Memory() != V4L2_MEMORY_MMAP) {
VLOGF(1) << "Called on invalid buffer type!";
return false;
}
return std::move(self).DoQueue(request_ref, nullptr);
}
bool V4L2WritableBufferRef::QueueUserPtr(const std::vector<void*>& ptrs,
V4L2RequestRef* request_ref) && {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK(buffer_data_);
// Move ourselves so our data gets freed no matter when we return
V4L2WritableBufferRef self(std::move(*this));
if (self.Memory() != V4L2_MEMORY_USERPTR) {
VLOGF(1) << "Called on invalid buffer type!";
return false;
}
if (ptrs.size() != self.PlanesCount()) {
VLOGF(1) << "Provided " << ptrs.size() << " pointers while we require "
<< self.buffer_data_->v4l2_buffer_.length << ".";
return false;
}
for (size_t i = 0; i < ptrs.size(); i++) {
self.buffer_data_->v4l2_buffer_.m.planes[i].m.userptr =
reinterpret_cast<unsigned long>(ptrs[i]);
}
return std::move(self).DoQueue(request_ref, nullptr);
}
bool V4L2WritableBufferRef::QueueDMABuf(const std::vector<base::ScopedFD>& fds,
V4L2RequestRef* request_ref) && {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK(buffer_data_);
// Move ourselves so our data gets freed no matter when we return
V4L2WritableBufferRef self(std::move(*this));
if (self.Memory() != V4L2_MEMORY_DMABUF) {
VLOGF(1) << "Called on invalid buffer type!";
return false;
}
if (!self.buffer_data_->CheckNumFDsForFormat(fds.size())) {
return false;
}
size_t num_planes = self.PlanesCount();
for (size_t i = 0; i < num_planes; i++) {
self.buffer_data_->v4l2_buffer_.m.planes[i].m.fd = fds[i].get();
}
return std::move(self).DoQueue(request_ref, nullptr);
}
bool V4L2WritableBufferRef::QueueDMABuf(scoped_refptr<FrameResource> frame,
V4L2RequestRef* request_ref) && {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK(buffer_data_);
// Move ourselves so our data gets freed no matter when we return
V4L2WritableBufferRef self(std::move(*this));
if (self.Memory() != V4L2_MEMORY_DMABUF) {
VLOGF(1) << "Called on invalid buffer type!";
return false;
}
// TODO(andrescj): consider replacing this by a DCHECK.
if (frame->storage_type() != VideoFrame::STORAGE_GPU_MEMORY_BUFFER &&
frame->storage_type() != VideoFrame::STORAGE_DMABUFS) {
VLOGF(1) << "Only frames with GpuMemoryBuffer and dma-buf are supported";
return false;
}
// The FDs duped by CreateGpuMemoryBufferHandle() will be closed after the
// call to DoQueue() which uses the VIDIOC_QBUF ioctl and so ends up
// increasing the reference count of the dma-buf. Thus, closing the FDs is
// safe.
// TODO(andrescj): for dma-buf frames, duping the FDs is unnecessary.
// Consider handling that path separately.
gfx::GpuMemoryBufferHandle gmb_handle = frame->CreateGpuMemoryBufferHandle();
if (gmb_handle.type != gfx::GpuMemoryBufferType::NATIVE_PIXMAP) {
VLOGF(1) << "Failed to create GpuMemoryBufferHandle for frame!";
return false;
}
const std::vector<gfx::NativePixmapPlane>& planes =
gmb_handle.native_pixmap_handle.planes;
if (!self.buffer_data_->CheckNumFDsForFormat(planes.size())) {
return false;
}
size_t num_planes = self.PlanesCount();
for (size_t i = 0; i < num_planes; i++) {
self.buffer_data_->v4l2_buffer_.m.planes[i].m.fd = planes[i].fd.get();
}
return std::move(self).DoQueue(request_ref, std::move(frame));
}
bool V4L2WritableBufferRef::QueueDMABuf(
const std::vector<gfx::NativePixmapPlane>& planes,
V4L2RequestRef* request_ref) && {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK(buffer_data_);
// Move ourselves so our data gets freed no matter when we return
V4L2WritableBufferRef self(std::move(*this));
if (self.Memory() != V4L2_MEMORY_DMABUF) {
VLOGF(1) << "Called on invalid buffer type!";
return false;
}
if (!self.buffer_data_->CheckNumFDsForFormat(planes.size())) {
return false;
}
size_t num_planes = self.PlanesCount();
for (size_t i = 0; i < num_planes; i++) {
self.buffer_data_->v4l2_buffer_.m.planes[i].m.fd = planes[i].fd.get();
}
return std::move(self).DoQueue(request_ref, nullptr);
}
bool V4L2WritableBufferRef::QueueDMABuf(uint64_t secure_handle,
V4L2RequestRef* request_ref) && {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK(buffer_data_);
// Move ourselves so our data gets freed no matter when we return
V4L2WritableBufferRef self(std::move(*this));
if (self.Memory() != V4L2_MEMORY_DMABUF) {
VLOGF(1) << "Called on invalid buffer type!";
return false;
}
// Set the FD for the secure handle.
bool set_fd = self.buffer_data_->queue_->SetBufferFdForSecureHandle(
secure_handle, &self.buffer_data_->v4l2_buffer_);
if (!set_fd) {
return false;
}
// The FD should already be set in the plane data, so submit it.
return std::move(self).DoQueue(request_ref, nullptr);
}
size_t V4L2WritableBufferRef::PlanesCount() const {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK(buffer_data_);
return buffer_data_->v4l2_buffer_.length;
}
size_t V4L2WritableBufferRef::GetPlaneSize(const size_t plane) const {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK(buffer_data_);
if (plane >= PlanesCount()) {
VLOGF(1) << "Invalid plane " << plane << " requested.";
return 0;
}
return buffer_data_->v4l2_buffer_.m.planes[plane].length;
}
void V4L2WritableBufferRef::SetPlaneSize(const size_t plane,
const size_t size) {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK(buffer_data_);
enum v4l2_memory memory = Memory();
if (memory == V4L2_MEMORY_MMAP) {
DCHECK_EQ(buffer_data_->v4l2_buffer_.m.planes[plane].length, size);
return;
}
DCHECK(memory == V4L2_MEMORY_USERPTR || memory == V4L2_MEMORY_DMABUF);
if (plane >= PlanesCount()) {
VLOGF(1) << "Invalid plane " << plane << " requested.";
return;
}
buffer_data_->v4l2_buffer_.m.planes[plane].length = size;
}
void* V4L2WritableBufferRef::GetPlaneMapping(const size_t plane) {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK(buffer_data_);
return buffer_data_->GetPlaneMapping(plane);
}
void V4L2WritableBufferRef::SetTimeStamp(const struct timeval& timestamp) {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK(buffer_data_);
buffer_data_->v4l2_buffer_.timestamp = timestamp;
}
const struct timeval& V4L2WritableBufferRef::GetTimeStamp() const {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK(buffer_data_);
return buffer_data_->v4l2_buffer_.timestamp;
}
void V4L2WritableBufferRef::SetPlaneBytesUsed(const size_t plane,
const size_t bytes_used) {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK(buffer_data_);
if (plane >= PlanesCount()) {
VLOGF(1) << "Invalid plane " << plane << " requested.";
return;
}
if (bytes_used > GetPlaneSize(plane)) {
VLOGF(1) << "Set bytes used " << bytes_used << " larger than plane size "
<< GetPlaneSize(plane) << ".";
return;
}
buffer_data_->v4l2_buffer_.m.planes[plane].bytesused = bytes_used;
}
size_t V4L2WritableBufferRef::GetPlaneBytesUsed(const size_t plane) const {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK(buffer_data_);
if (plane >= PlanesCount()) {
VLOGF(1) << "Invalid plane " << plane << " requested.";
return 0;
}
return buffer_data_->v4l2_buffer_.m.planes[plane].bytesused;
}
void V4L2WritableBufferRef::SetPlaneDataOffset(const size_t plane,
const size_t data_offset) {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK(buffer_data_);
if (plane >= PlanesCount()) {
VLOGF(1) << "Invalid plane " << plane << " requested.";
return;
}
buffer_data_->v4l2_buffer_.m.planes[plane].data_offset = data_offset;
}
size_t V4L2WritableBufferRef::BufferId() const {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK(buffer_data_);
return buffer_data_->v4l2_buffer_.index;
}
V4L2ReadableBuffer::V4L2ReadableBuffer(const struct v4l2_buffer& v4l2_buffer,
base::WeakPtr<V4L2Queue> queue,
scoped_refptr<FrameResource> frame)
: buffer_data_(
std::make_unique<V4L2BufferRefBase>(v4l2_buffer, std::move(queue))),
frame_(std::move(frame)) {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
}
scoped_refptr<FrameResource> V4L2ReadableBuffer::GetFrameResource() {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK(buffer_data_);
if (buffer_data_->v4l2_buffer_.memory == V4L2_MEMORY_DMABUF && frame_) {
return frame_;
}
return buffer_data_->GetFrameResource();
}
V4L2ReadableBuffer::~V4L2ReadableBuffer() {
// This method is thread-safe. Since we are the destructor, we are guaranteed
// to be called from the only remaining reference to us. Also, we are just
// calling the destructor of buffer_data_, which is also thread-safe.
DCHECK(buffer_data_);
}
bool V4L2ReadableBuffer::IsLast() const {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK(buffer_data_);
return buffer_data_->v4l2_buffer_.flags & V4L2_BUF_FLAG_LAST;
}
bool V4L2ReadableBuffer::IsKeyframe() const {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK(buffer_data_);
return buffer_data_->v4l2_buffer_.flags & V4L2_BUF_FLAG_KEYFRAME;
}
bool V4L2ReadableBuffer::IsError() const {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK(buffer_data_);
// "The driver may also set V4L2_BUF_FLAG_ERROR in the flags field. It
// indicates a non-critical (recoverable) streaming error. In such case the
// application may continue as normal, but should be aware that data in the
// dequeued buffer might be corrupted." IOW it is more a discard-this-buffer
// marker than a fatal error indication, so it's down to the caller to take
// action if needed/desired.
// https://www.kernel.org/doc/html/v5.15/userspace-api/media/v4l/vidioc-qbuf.html#description
return buffer_data_->v4l2_buffer_.flags & V4L2_BUF_FLAG_ERROR;
}
struct timeval V4L2ReadableBuffer::GetTimeStamp() const {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK(buffer_data_);
return buffer_data_->v4l2_buffer_.timestamp;
}
size_t V4L2ReadableBuffer::PlanesCount() const {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK(buffer_data_);
return buffer_data_->v4l2_buffer_.length;
}
const void* V4L2ReadableBuffer::GetPlaneMapping(const size_t plane) const {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK(buffer_data_);
return buffer_data_->GetPlaneMapping(plane);
}
size_t V4L2ReadableBuffer::GetPlaneBytesUsed(const size_t plane) const {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK(buffer_data_);
if (plane >= PlanesCount()) {
VLOGF(1) << "Invalid plane " << plane << " requested.";
return 0;
}
return buffer_data_->v4l2_planes_[plane].bytesused;
}
size_t V4L2ReadableBuffer::GetPlaneDataOffset(const size_t plane) const {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK(buffer_data_);
if (plane >= PlanesCount()) {
VLOGF(1) << "Invalid plane " << plane << " requested.";
return 0;
}
return buffer_data_->v4l2_planes_[plane].data_offset;
}
size_t V4L2ReadableBuffer::BufferId() const {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK(buffer_data_);
return buffer_data_->v4l2_buffer_.index;
}
struct SecureBufferData {
SecureBufferData(uint64_t in_secure_handle, base::ScopedFD in_fd)
: secure_handle(in_secure_handle), fd(std::move(in_fd)) {}
SecureBufferData(SecureBufferData&& other) = default;
~SecureBufferData() {}
// true if the secure buffer stores decrypted data from an active
// DecoderBuffer.
bool owned_by_decoder_buffer = false;
// List of all the buffer indexes that have this FD/secure_handle currently
// attached to it.
std::vector<size_t> queued_buffer_indexes;
uint64_t secure_handle;
base::ScopedFD fd;
};
// Helper macros that print the queue type with logs.
#define VPQLOGF(level) \
VPLOGF(level) << "(" << V4L2BufferTypeToString(type_) << ") "
#define VQLOGF(level) \
VLOGF(level) << "(" << V4L2BufferTypeToString(type_) << ") "
#define DVQLOGF(level) \
DVLOGF(level) << "(" << V4L2BufferTypeToString(type_) << ") "
V4L2Queue::V4L2Queue(const IoctlAsCallback& ioctl_cb,
const base::RepeatingClosure& schedule_poll_cb,
const MmapAsCallback& mmap_cb,
const AllocateSecureBufferAsCallback& allocate_secure_cb,
enum v4l2_buf_type type,
base::OnceClosure destroy_cb)
: type_(type),
ioctl_cb_(ioctl_cb),
schedule_poll_cb_(schedule_poll_cb),
mmap_cb_(mmap_cb),
allocate_secure_cb_(allocate_secure_cb),
destroy_cb_(std::move(destroy_cb)),
weak_this_factory_(this) {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
struct v4l2_requestbuffers reqbufs = {
.count = 0, .type = type_, .memory = V4L2_MEMORY_MMAP};
supports_requests_ = (ioctl_cb_.Run(VIDIOC_REQBUFS, &reqbufs) == kIoctlOk) &&
(reqbufs.capabilities & V4L2_BUF_CAP_SUPPORTS_REQUESTS);
// Stateful backends for example do not support requests.
VPLOG_IF(4, supports_requests_)
<< "This queue does " << (supports_requests_ ? "" : "not")
<< " support requests.";
}
V4L2Queue::~V4L2Queue() {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
if (is_streaming_ && !Streamoff()) {
VQLOGF(1) << "Failed to stop queue";
}
DCHECK(queued_buffers_.empty());
if (!buffers_.empty() && !DeallocateBuffers()) {
VQLOGF(1) << "Failed to deallocate queue buffers";
}
std::move(destroy_cb_).Run();
}
std::optional<struct v4l2_format> V4L2Queue::SetFormat(uint32_t fourcc,
const gfx::Size& size,
size_t buffer_size) {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
struct v4l2_format format = BuildV4L2Format(type_, fourcc, size, buffer_size);
if (ioctl_cb_.Run(VIDIOC_S_FMT, &format) != 0 ||
format.fmt.pix_mp.pixelformat != fourcc) {
RecordVidiocIoctlErrorUMA(VidiocIoctlRequests::kVidiocSFmt);
VPQLOGF(2) << "Failed to set format fourcc: " << FourccToString(fourcc);
return std::nullopt;
}
current_format_ = format;
return current_format_;
}
std::optional<struct v4l2_format> V4L2Queue::TryFormat(uint32_t fourcc,
const gfx::Size& size,
size_t buffer_size) {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
struct v4l2_format format = BuildV4L2Format(type_, fourcc, size, buffer_size);
if (ioctl_cb_.Run(VIDIOC_TRY_FMT, &format) != 0 ||
format.fmt.pix_mp.pixelformat != fourcc) {
VPQLOGF(2) << "Failed to try format fourcc: " << FourccToString(fourcc);
return std::nullopt;
}
return format;
}
std::pair<std::optional<struct v4l2_format>, int> V4L2Queue::GetFormat() {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
struct v4l2_format format;
memset(&format, 0, sizeof(format));
format.type = type_;
if (ioctl_cb_.Run(VIDIOC_G_FMT, &format) != 0) {
RecordVidiocIoctlErrorUMA(VidiocIoctlRequests::kVidiocGFmt);
VPQLOGF(2) << "Failed to get format";
return std::make_pair(std::nullopt, errno);
}
return std::make_pair(format, 0);
}
std::optional<gfx::Rect> V4L2Queue::GetVisibleRect() {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
struct v4l2_selection selection = {.type = type_,
.target = V4L2_SEL_TGT_COMPOSE};
if (ioctl_cb_.Run(VIDIOC_G_SELECTION, &selection) != 0) {
RecordVidiocIoctlErrorUMA(VidiocIoctlRequests::kVidiocGSelection);
VQLOGF(1) << "Failed to get visible rect";
return std::nullopt;
}
return V4L2RectToGfxRect(selection.r);
}
size_t V4L2Queue::AllocateBuffers(size_t count,
enum v4l2_memory memory,
bool incoherent) {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK(!free_buffers_);
DCHECK(queued_buffers_.empty());
incoherent_ = incoherent;
if (IsStreaming()) {
VQLOGF(1) << "Cannot allocate buffers while streaming.";
return 0;
}
if (buffers_.size() != 0) {
VQLOGF(1)
<< "Cannot allocate new buffers while others are still allocated.";
return 0;
}
// Should have been cleared in DeallocateBuffers() if it was ever filled in.
DCHECK(free_buffers_indexes_.empty());
if (count == 0) {
VQLOGF(1) << "Attempting to allocate 0 buffers.";
return 0;
}
// First query the number of planes in the buffers we are about to request.
std::optional<v4l2_format> format = GetFormat().first;
if (!format) {
VQLOGF(1) << "Cannot get format.";
return 0;
}
planes_count_ = format->fmt.pix_mp.num_planes;
DCHECK_LE(planes_count_, static_cast<size_t>(VIDEO_MAX_PLANES));
__u8 flags = incoherent ? V4L2_MEMORY_FLAG_NON_COHERENT : 0;
if (allocate_secure_cb_) {
flags |= V4L2_MEMORY_FLAG_SECURE;
}
struct v4l2_requestbuffers reqbufs = {
.count = base::checked_cast<decltype(v4l2_requestbuffers::count)>(count),
.type = type_,
.memory = memory,
.flags = flags};
DVQLOGF(3) << "Requesting " << count << " buffers ("
<< (incoherent ? "incoherent" : "coherent") << ")";
int ret = ioctl_cb_.Run(VIDIOC_REQBUFS, &reqbufs);
if (ret) {
RecordVidiocIoctlErrorUMA(VidiocIoctlRequests::kVidiocReqbufs);
VPQLOGF(1) << "VIDIOC_REQBUFS failed";
return 0;
}
DVQLOGF(3) << "Allocated " << reqbufs.count << " buffers.";
memory_ = memory;
free_buffers_ = new V4L2BuffersList();
// Now query all buffer information.
for (size_t i = 0; i < reqbufs.count; i++) {
auto buffer =
V4L2Buffer::Create(ioctl_cb_, mmap_cb_, type_, memory_, *format, i);
if (!buffer) {
if (!DeallocateBuffers()) {
VQLOGF(1) << "Failed to deallocate queue buffers";
}
return 0;
}
if (type_ == V4L2_BUF_TYPE_VIDEO_OUTPUT_MPLANE && allocate_secure_cb_) {
CHECK_EQ(memory_, V4L2_MEMORY_DMABUF);
// Invoke the callback for secure buffer allocation. We only use dmabufs
// for the OUTPUT queue when doing secure playback.
allocate_secure_cb_.Run(buffer->v4l2_buffer().m.planes[0].length,
base::BindOnce(&V4L2Queue::SecureBufferAllocated,
weak_this_factory_.GetWeakPtr()));
}
buffers_.emplace_back(std::move(buffer));
free_buffers_->ReturnBuffer(i);
}
DCHECK(free_buffers_);
DCHECK_EQ(free_buffers_->size(), buffers_.size());
DCHECK(queued_buffers_.empty());
return buffers_.size();
}
bool V4L2Queue::DeallocateBuffers() {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
if (IsStreaming()) {
VQLOGF(1) << "Cannot deallocate buffers while streaming.";
return false;
}
if (buffers_.size() == 0) {
return true;
}
weak_this_factory_.InvalidateWeakPtrs();
buffers_.clear();
free_buffers_indexes_.clear();
free_buffers_ = nullptr;
secure_buffers_.clear();
// Free all buffers.
__u8 flags = incoherent_ ? V4L2_MEMORY_FLAG_NON_COHERENT : 0;
if (allocate_secure_cb_) {
flags |= V4L2_MEMORY_FLAG_SECURE;
}
struct v4l2_requestbuffers reqbufs = {
.count = 0, .type = type_, .memory = memory_, .flags = flags};
int ret = ioctl_cb_.Run(VIDIOC_REQBUFS, &reqbufs);
if (ret) {
RecordVidiocIoctlErrorUMA(VidiocIoctlRequests::kVidiocReqbufs);
VPQLOGF(1) << "VIDIOC_REQBUFS failed";
return false;
}
DCHECK(!free_buffers_);
DCHECK(queued_buffers_.empty());
return true;
}
size_t V4L2Queue::GetMemoryUsage() const {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
size_t usage = 0;
for (const auto& buf : buffers_) {
usage += buf->GetMemoryUsage();
}
return usage;
}
v4l2_memory V4L2Queue::GetMemoryType() const {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
return memory_;
}
// This class is used to expose buffer reference classes constructors to
// this module. This is to ensure that nobody else can create buffer references.
class V4L2BufferRefFactory {
public:
static V4L2WritableBufferRef CreateWritableRef(
const struct v4l2_buffer& v4l2_buffer,
base::WeakPtr<V4L2Queue> queue) {
return V4L2WritableBufferRef(v4l2_buffer, std::move(queue));
}
static V4L2ReadableBufferRef CreateReadableRef(
const struct v4l2_buffer& v4l2_buffer,
base::WeakPtr<V4L2Queue> queue,
scoped_refptr<FrameResource> frame) {
return new V4L2ReadableBuffer(v4l2_buffer, std::move(queue),
std::move(frame));
}
};
CroStatus::Or<uint64_t> V4L2Queue::GetFreeSecureHandle() {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
// Go through the list of secure buffers and find one that is not owned or
// queued.
for (auto& buf : secure_buffers_) {
if (!buf.owned_by_decoder_buffer && buf.queued_buffer_indexes.empty()) {
buf.owned_by_decoder_buffer = true;
return buf.secure_handle;
}
}
return CroStatus::Codes::kSecureBufferPoolEmpty;
}
void V4L2Queue::ReleaseSecureHandle(uint64_t secure_handle) {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
// Find the matching secure buffer and release the ownership on it, it might
// still be in use, but that would be tracked by the queue counter if so.
for (auto& buf : secure_buffers_) {
if (buf.secure_handle == secure_handle) {
buf.owned_by_decoder_buffer = false;
return;
}
}
}
std::optional<V4L2WritableBufferRef> V4L2Queue::GetFreeBuffer() {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
// No buffers allocated at the moment?
if (!free_buffers_) {
return std::nullopt;
}
auto buffer_id = free_buffers_->GetFreeBuffer();
if (!buffer_id.has_value()) {
return std::nullopt;
}
return V4L2BufferRefFactory::CreateWritableRef(
buffers_[buffer_id.value()]->v4l2_buffer(),
weak_this_factory_.GetWeakPtr());
}
std::optional<V4L2WritableBufferRef> V4L2Queue::GetFreeBuffer(
size_t requested_buffer_id) {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
// No buffers allocated at the moment?
if (!free_buffers_) {
return std::nullopt;
}
auto buffer_id = free_buffers_->GetFreeBuffer(requested_buffer_id);
if (!buffer_id.has_value()) {
return std::nullopt;
}
return V4L2BufferRefFactory::CreateWritableRef(
buffers_[buffer_id.value()]->v4l2_buffer(),
weak_this_factory_.GetWeakPtr());
}
std::optional<V4L2WritableBufferRef> V4L2Queue::GetFreeBufferForFrame(
const gfx::GenericSharedMemoryId& id) {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
// No buffers allocated at the moment?
if (!free_buffers_) {
return std::nullopt;
}
if (memory_ != V4L2_MEMORY_DMABUF) {
DVLOGF(1) << "Queue is not DMABUF";
return std::nullopt;
}
if (!id.is_valid()) {
DVLOGF(1) << "Provided identifier was not valid";
return std::nullopt;
}
// If |id| has already been used in |buffers_|, then return that buffer.
// Otherwise use the next buffer from |free_buffers_indexes_|.
if (!base::Contains(free_buffers_indexes_, id)) {
if (free_buffers_indexes_.size() >= buffers_.size()) {
return std::nullopt;
}
// The value for |id| is simply the map size(): a poor man's way to have a
// monotonically increasing counter.
free_buffers_indexes_.emplace(id, free_buffers_indexes_.size());
}
return GetFreeBuffer(free_buffers_indexes_[id]);
}
bool V4L2Queue::QueueBuffer(struct v4l2_buffer* v4l2_buffer,
scoped_refptr<FrameResource> frame) {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
V4L2ProcessingTrace(v4l2_buffer, /*start=*/true);
int ret = ioctl_cb_.Run(VIDIOC_QBUF, v4l2_buffer);
if (ret) {
RecordVidiocIoctlErrorUMA(VidiocIoctlRequests::kVidiocQbuf);
VPQLOGF(1) << "VIDIOC_QBUF failed";
return false;
}
const auto inserted =
queued_buffers_.emplace(v4l2_buffer->index, std::move(frame));
DCHECK(inserted.second);
schedule_poll_cb_.Run();
return true;
}
std::pair<bool, V4L2ReadableBufferRef> V4L2Queue::DequeueBuffer() {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
// No need to dequeue if no buffers queued.
if (QueuedBuffersCount() == 0) {
return std::make_pair(true, nullptr);
}
if (!IsStreaming()) {
VQLOGF(1) << "Attempting to dequeue a buffer while not streaming.";
return std::make_pair(true, nullptr);
}
struct v4l2_buffer v4l2_buffer;
memset(&v4l2_buffer, 0, sizeof(v4l2_buffer));
// WARNING: do not change this to a vector or something smaller than
// VIDEO_MAX_PLANES (the maximum number of planes V4L2 supports). The
// element overhead is small and may avoid memory corruption bugs.
struct v4l2_plane planes[VIDEO_MAX_PLANES];
memset(planes, 0, sizeof(planes));
v4l2_buffer.type = type_;
v4l2_buffer.memory = memory_;
v4l2_buffer.m.planes = planes;
v4l2_buffer.length = planes_count_;
int ret = ioctl_cb_.Run(VIDIOC_DQBUF, &v4l2_buffer);
if (ret) {
// TODO(acourbot): we should not have to check for EPIPE as codec clients
// should not call this method after the last buffer is dequeued.
switch (errno) {
case EAGAIN:
case EPIPE:
// This is not an error so we'll need to continue polling but won't
// provide a buffer.
schedule_poll_cb_.Run();
return std::make_pair(true, nullptr);
default:
RecordVidiocIoctlErrorUMA(VidiocIoctlRequests::kVidiocDqbuf);
VPQLOGF(1) << "VIDIOC_DQBUF failed";
return std::make_pair(false, nullptr);
}
}
auto it = queued_buffers_.find(v4l2_buffer.index);
CHECK(it != queued_buffers_.end(), base::NotFatalUntil::M130);
scoped_refptr<FrameResource> queued_frame = std::move(it->second);
queued_buffers_.erase(it);
V4L2ProcessingTrace(&v4l2_buffer, /*start=*/false);
if (QueuedBuffersCount() > 0) {
schedule_poll_cb_.Run();
}
// See if we need to remove this from any of the secure buffer queue tracking.
for (auto& buf : secure_buffers_) {
std::erase_if(buf.queued_buffer_indexes, [v4l2_buffer](size_t idx) {
return idx == v4l2_buffer.index;
});
}
DCHECK(free_buffers_);
return std::make_pair(true, V4L2BufferRefFactory::CreateReadableRef(
v4l2_buffer, weak_this_factory_.GetWeakPtr(),
std::move(queued_frame)));
}
bool V4L2Queue::IsStreaming() const {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
return is_streaming_;
}
bool V4L2Queue::Streamon() {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
if (is_streaming_) {
return true;
}
int arg = static_cast<int>(type_);
int ret = ioctl_cb_.Run(VIDIOC_STREAMON, &arg);
if (ret) {
RecordVidiocIoctlErrorUMA(VidiocIoctlRequests::kVidiocStreamon);
VPQLOGF(1) << "VIDIOC_STREAMON failed";
return false;
}
is_streaming_ = true;
return true;
}
bool V4L2Queue::Streamoff() {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
// We do not check the value of IsStreaming(), because we may have queued
// buffers to the queue and wish to get them back - in such as case, we may
// need to do a VIDIOC_STREAMOFF on a stopped queue.
int arg = static_cast<int>(type_);
int ret = ioctl_cb_.Run(VIDIOC_STREAMOFF, &arg);
if (ret) {
RecordVidiocIoctlErrorUMA(VidiocIoctlRequests::kVidiocStreamoff);
VPQLOGF(1) << "VIDIOC_STREAMOFF failed";
return false;
}
for (const auto& it : queued_buffers_) {
DCHECK(free_buffers_);
free_buffers_->ReturnBuffer(it.first);
}
for (auto& buf : secure_buffers_) {
buf.queued_buffer_indexes.clear();
}
queued_buffers_.clear();
is_streaming_ = false;
return true;
}
size_t V4L2Queue::AllocatedBuffersCount() const {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
return buffers_.size();
}
size_t V4L2Queue::FreeBuffersCount() const {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
return free_buffers_ ? free_buffers_->size() : 0;
}
size_t V4L2Queue::QueuedBuffersCount() const {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
return queued_buffers_.size();
}
#undef VDQLOGF
#undef VPQLOGF
#undef VQLOGF
bool V4L2Queue::SupportsRequests() {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
return supports_requests_;
}
std::optional<struct v4l2_format> V4L2Queue::SetModifierFormat(
uint64_t modifier,
const gfx::Size& size) {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
if (DRM_FORMAT_MOD_QCOM_COMPRESSED == modifier) {
auto format = SetFormat(V4L2_PIX_FMT_QC08C, size, 0);
if (!format) {
VPLOGF(1) << "Failed to set magic modifier format.";
}
return format;
}
return std::nullopt;
}
bool V4L2Queue::SendStopCommand() {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
return SendCommand(V4L2_DEC_CMD_STOP);
}
bool V4L2Queue::SendStartCommand() {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
return SendCommand(V4L2_DEC_CMD_START);
}
bool V4L2Queue::SetBufferFdForSecureHandle(uint64_t secure_handle,
struct v4l2_buffer* v4l2_buffer) {
for (auto& buf : secure_buffers_) {
if (buf.secure_handle == secure_handle) {
if (!buf.owned_by_decoder_buffer) {
return false;
}
buf.queued_buffer_indexes.emplace_back(v4l2_buffer->index);
v4l2_buffer->m.planes[0].m.fd = buf.fd.get();
return true;
}
}
return false;
}
bool V4L2Queue::SendCommand(__u32 command) {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
// TODO(mcasas): Restrict this to V4L2_BUF_TYPE_VIDEO_OUTPUT_MPLANE, after
// deprecating V4L2StatefulVideoDecoderBackend.
struct v4l2_decoder_cmd cmd;
memset(&cmd, 0, sizeof(cmd)); // Must use memset() due to unions.
cmd.cmd = command;
const bool success = ioctl_cb_.Run(VIDIOC_DECODER_CMD, &cmd) == kIoctlOk;
PLOG_IF(ERROR, !success) << "Failed to issue command " << command
<< " (V4L2_DEC_CMD_START: " << V4L2_DEC_CMD_START
<< ", V4L2_DEC_CMD_STOP: " << V4L2_DEC_CMD_STOP
<< ")";
return success;
}
void V4L2Queue::SecureBufferAllocated(base::ScopedFD secure_fd,
uint64_t secure_handle) {
CHECK(secure_fd.is_valid());
CHECK(secure_handle);
secure_buffers_.emplace_back(secure_handle, std::move(secure_fd));
}
class V4L2Request {
public:
V4L2Request(const V4L2Request&) = delete;
V4L2Request& operator=(const V4L2Request&) = delete;
// Apply the passed controls to the request.
bool ApplyCtrls(struct v4l2_ext_controls* ctrls);
// Apply the passed buffer to the request..
bool ApplyQueueBuffer(struct v4l2_buffer* buffer);
// Submits the request to the driver.
bool Submit();
// Indicates if the request has completed.
bool IsCompleted();
// Waits for the request to complete for a determined timeout. Returns false
// if the request is not ready or other error. Default timeout is 500ms.
bool WaitForCompletion(int poll_timeout_ms = 500);
// Resets the request.
bool Reset();
private:
raw_ptr<V4L2RequestsQueue> request_queue_;
int ref_counter_ = 0;
base::ScopedFD request_fd_;
friend class V4L2RequestsQueue;
V4L2Request(base::ScopedFD&& request_fd, V4L2RequestsQueue* request_queue)
: request_queue_(request_queue), request_fd_(std::move(request_fd)) {}
friend class V4L2RequestRefBase;
// Increases the number of request references.
void IncRefCounter();
// Decreases the number of request references.
// When the counters reaches zero, the request is returned to the queue.
int DecRefCounter();
SEQUENCE_CHECKER(sequence_checker_);
};
void V4L2Request::IncRefCounter() {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
ref_counter_++;
}
int V4L2Request::DecRefCounter() {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
ref_counter_--;
if (ref_counter_ < 1) {
request_queue_->ReturnRequest(this);
}
return ref_counter_;
}
bool V4L2Request::ApplyCtrls(struct v4l2_ext_controls* ctrls) {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK_NE(ctrls, nullptr);
if (!request_fd_.is_valid()) {
VPLOGF(1) << "Invalid request";
return false;
}
ctrls->which = V4L2_CTRL_WHICH_REQUEST_VAL;
ctrls->request_fd = request_fd_.get();
return true;
}
bool V4L2Request::ApplyQueueBuffer(struct v4l2_buffer* buffer) {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK_NE(buffer, nullptr);
if (!request_fd_.is_valid()) {
VPLOGF(1) << "Invalid request";
return false;
}
buffer->flags |= V4L2_BUF_FLAG_REQUEST_FD;
buffer->request_fd = request_fd_.get();
return true;
}
bool V4L2Request::Submit() {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
if (!request_fd_.is_valid()) {
VPLOGF(1) << "No valid request file descriptor to submit request.";
return false;
}
if (HANDLE_EINTR(ioctl(request_fd_.get(), MEDIA_REQUEST_IOC_QUEUE)) != 0) {
RecordMediaIoctlUMA(MediaIoctlRequests::kMediaRequestIocQueue);
return false;
}
return true;
}
bool V4L2Request::IsCompleted() {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
return WaitForCompletion(0);
}
bool V4L2Request::WaitForCompletion(int poll_timeout_ms) {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
if (!request_fd_.is_valid()) {
VPLOGF(1) << "Invalid request";
return false;
}
struct pollfd poll_fd = {request_fd_.get(), POLLPRI, 0};
// Poll the request to ensure its previous task is done
switch (poll(&poll_fd, 1, poll_timeout_ms)) {
case 1:
return true;
case 0:
// Not an error - we just timed out.
DVLOGF(4) << "Request poll(" << poll_timeout_ms << ") timed out";
return false;
case -1:
VPLOGF(1) << "Failed to poll request";
return false;
default:
NOTREACHED();
}
}
bool V4L2Request::Reset() {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
if (!request_fd_.is_valid()) {
VPLOGF(1) << "Invalid request";
return false;
}
// Reinit the request to make sure we can use it for a new submission.
if (HANDLE_EINTR(ioctl(request_fd_.get(), MEDIA_REQUEST_IOC_REINIT)) < 0) {
RecordMediaIoctlUMA(MediaIoctlRequests::kMediaRequestIocReinit);
VPLOGF(1) << "Failed to reinit request.";
return false;
}
return true;
}
V4L2RequestRefBase::V4L2RequestRefBase(V4L2RequestRefBase&& req_base) {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
request_ = req_base.request_;
req_base.request_ = nullptr;
}
V4L2RequestRefBase::V4L2RequestRefBase(V4L2Request* request) {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
if (request) {
request_ = request;
request_->IncRefCounter();
}
}
V4L2RequestRefBase::~V4L2RequestRefBase() {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
if (request_) {
request_->DecRefCounter();
}
}
bool V4L2RequestRef::ApplyCtrls(struct v4l2_ext_controls* ctrls) const {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK_NE(request_, nullptr);
return request_->ApplyCtrls(ctrls);
}
bool V4L2RequestRef::ApplyQueueBuffer(struct v4l2_buffer* buffer) const {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK_NE(request_, nullptr);
return request_->ApplyQueueBuffer(buffer);
}
std::optional<V4L2SubmittedRequestRef> V4L2RequestRef::Submit() && {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK_NE(request_, nullptr);
V4L2RequestRef self(std::move(*this));
if (!self.request_->Submit()) {
return std::nullopt;
}
return V4L2SubmittedRequestRef(self.request_);
}
bool V4L2SubmittedRequestRef::IsCompleted() {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK_NE(request_, nullptr);
return request_->IsCompleted();
}
V4L2RequestsQueue::V4L2RequestsQueue(base::ScopedFD&& media_fd) {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
media_fd_ = std::move(media_fd);
}
V4L2RequestsQueue::~V4L2RequestsQueue() {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
requests_.clear();
media_fd_.reset();
}
std::optional<base::ScopedFD> V4L2RequestsQueue::CreateRequestFD() {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
int request_fd;
int ret = HANDLE_EINTR(
ioctl(media_fd_.get(), MEDIA_IOC_REQUEST_ALLOC, &request_fd));
if (ret < 0) {
RecordMediaIoctlUMA(MediaIoctlRequests::kMediaIocRequestAlloc);
VPLOGF(1) << "Failed to create request";
return std::nullopt;
}
return base::ScopedFD(request_fd);
}
std::optional<V4L2RequestRef> V4L2RequestsQueue::GetFreeRequest() {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
V4L2Request* request_ptr =
free_requests_.empty() ? nullptr : free_requests_.front();
if (request_ptr && request_ptr->IsCompleted()) {
// Previous request is already completed, just recycle it.
free_requests_.pop();
} else if (requests_.size() < kMaxNumRequests) {
// No request yet, or not completed, but we can allocate a new one.
auto request_fd = CreateRequestFD();
if (!request_fd.has_value()) {
VLOGF(1) << "Error while creating a new request FD!";
return std::nullopt;
}
// Not using std::make_unique because constructor is private.
std::unique_ptr<V4L2Request> request(
new V4L2Request(std::move(*request_fd), this));
request_ptr = request.get();
requests_.push_back(std::move(request));
VLOGF(4) << "Allocated new request, total number: " << requests_.size();
} else {
// Request is not completed and we have reached the maximum number.
// Wait for it to complete.
VLOGF(1) << "Waiting for request completion. This probably means a "
<< "request is blocking.";
if (!request_ptr->WaitForCompletion()) {
VLOG(1) << "Timeout while waiting for request to complete.";
return std::nullopt;
}
free_requests_.pop();
}
DCHECK(request_ptr);
if (!request_ptr->Reset()) {
VPLOGF(1) << "Failed to reset request";
return std::nullopt;
}
return V4L2RequestRef(request_ptr);
}
void V4L2RequestsQueue::ReturnRequest(V4L2Request* request) {
DCHECK_CALLED_ON_VALID_SEQUENCE(sequence_checker_);
DCHECK(request);
if (request) {
free_requests_.push(request);
}
}
} // namespace media