1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633
|
// Copyright 2018 The Chromium Authors
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.
#include "gpu/command_buffer/service/image_reader_gl_owner.h"
#include <android/native_window_jni.h>
#include <jni.h>
#include <stdint.h>
#include "base/android/android_hardware_buffer_compat.h"
#include "base/android/build_info.h"
#include "base/android/jni_android.h"
#include "base/android/scoped_hardware_buffer_fence_sync.h"
#include "base/debug/dump_without_crashing.h"
#include "base/logging.h"
#include "base/memory/ptr_util.h"
#include "base/memory/raw_ptr.h"
#include "base/metrics/histogram_functions.h"
#include "base/metrics/histogram_macros.h"
#include "base/posix/eintr_wrapper.h"
#include "base/strings/stringprintf.h"
#include "base/synchronization/waitable_event.h"
#include "base/task/bind_post_task.h"
#include "gpu/command_buffer/service/abstract_texture_android.h"
#include "gpu/config/gpu_finch_features.h"
#include "ui/gfx/android/android_surface_control_compat.h"
#include "ui/gl/gl_fence_android_native_fence_sync.h"
#include "ui/gl/gl_utils.h"
#include "ui/gl/scoped_binders.h"
#include "ui/gl/scoped_make_current.h"
namespace gpu {
namespace {
BASE_FEATURE(kDiscardDroppedEarlyRenderedFrames,
"DiscardDroppedEarlyRenderedFrames",
base::FEATURE_ENABLED_BY_DEFAULT);
bool IsSurfaceControl(TextureOwner::Mode mode) {
switch (mode) {
case TextureOwner::Mode::kAImageReaderInsecureSurfaceControl:
case TextureOwner::Mode::kAImageReaderSecureSurfaceControl:
return true;
case TextureOwner::Mode::kAImageReaderInsecure:
return false;
case TextureOwner::Mode::kSurfaceTextureInsecure:
NOTREACHED();
}
NOTREACHED();
}
// This should be as small as possible to limit the memory usage.
// ImageReader needs 1 image to mimic the behavior of SurfaceTexture but
// 2 images are required to minimize negative impact on
// smoothness. This is because in case an image is not acquired for some
// reasons, last acquired image should be displayed which is only possible with
// 2 images (1 previously acquired, 1 currently acquired/tried to acquire).
// But some devices supports only 1 image to be acquired. (see
// crbug.com/1051705). For SurfaceControl we need 3 images instead of 2 since 1
// frame (and hence image associated with it) will be with system compositor and
// 2 frames will be in flight. For multi-threaded compositor, when AImageReader
// is supported, we need 3 images in order to skip texture copy. 1 frame with
// display compositor, 1 frame in flight and 1 frame being prepared by the
// renderer.
uint32_t NumRequiredMaxImages(TextureOwner::Mode mode) {
if (IsSurfaceControl(mode)) {
DCHECK(!features::LimitAImageReaderMaxSizeToOne());
if (features::IncreaseBufferCountForHighFrameRate())
return 5;
return 3;
}
return features::LimitAImageReaderMaxSizeToOne() ? 1 : 2;
}
std::optional<gfx::Size> GetImageSize(AImage* image) {
int32_t width = 0, height = 0;
if (AImage_getWidth(image, &width) != AMEDIA_OK ||
AImage_getHeight(image, &height) != AMEDIA_OK || width <= 0 ||
height <= 0) {
return std::nullopt;
}
return gfx::Size(width, height);
}
} // namespace
// This class is safe to be created/destroyed on different threads. This is made
// sure by destruction happening on correct thread. This class is not thread
// safe to be used concurrently on multiple thraeads.
class ImageReaderGLOwner::ScopedHardwareBufferImpl
: public base::android::ScopedHardwareBufferFenceSync {
public:
ScopedHardwareBufferImpl(scoped_refptr<ImageReaderGLOwner> texture_owner,
AImage* image,
base::android::ScopedHardwareBufferHandle handle,
base::ScopedFD fence_fd)
: base::android::ScopedHardwareBufferFenceSync(std::move(handle),
std::move(fence_fd),
base::ScopedFD()),
texture_owner_(std::move(texture_owner)),
image_(image) {
DCHECK(image_);
texture_owner_->RegisterRefOnImageLocked(image_);
}
~ScopedHardwareBufferImpl() override {
texture_owner_->ReleaseRefOnImage(image_, std::move(read_fence_));
}
void SetReadFence(base::ScopedFD fence_fd) final {
// Client can call this method multiple times for a hardware buffer. Hence
// all the client provided sync_fd should be merged. Eg: BeginReadAccess()
// can be called multiple times for an AndroidVideoImageBacking
// representation.
read_fence_ = gl::MergeFDs(std::move(read_fence_), std::move(fence_fd));
}
private:
base::ScopedFD read_fence_;
scoped_refptr<ImageReaderGLOwner> texture_owner_;
raw_ptr<AImage> image_;
};
ImageReaderGLOwner::ImageReaderGLOwner(
std::unique_ptr<AbstractTextureAndroid> texture,
Mode mode,
scoped_refptr<SharedContextState> context_state,
scoped_refptr<RefCountedLock> drdc_lock,
TextureOwnerCodecType type_for_metrics)
: TextureOwner(false /* binds_texture_on_image_update */,
std::move(texture),
std::move(context_state)),
RefCountedLockHelperDrDc(std::move(drdc_lock)),
context_(gl::GLContext::GetCurrent()),
surface_(gl::GLSurface::GetCurrent()),
type_for_metrics_(type_for_metrics) {
DCHECK(context_);
DCHECK(surface_);
// Set the width, height and format to some default value. This parameters
// are/maybe overriden by the producer sending buffers to this imageReader's
// Surface.
int32_t width = 1, height = 1;
max_images_ = NumRequiredMaxImages(mode);
AImageReader* reader = nullptr;
// The usage flag below should be used when the buffer will be read from by
// the GPU as a texture.
uint64_t usage = mode == Mode::kAImageReaderSecureSurfaceControl
? AHARDWAREBUFFER_USAGE_PROTECTED_CONTENT
: AHARDWAREBUFFER_USAGE_GPU_SAMPLED_IMAGE;
if (IsSurfaceControl(mode))
usage |= AHARDWAREBUFFER_USAGE_COMPOSER_OVERLAY;
// Create a new reader for images of the desired size and format.
media_status_t return_code = AImageReader_newWithUsage(
width, height, AIMAGE_FORMAT_PRIVATE, usage, max_images_, &reader);
if (return_code != AMEDIA_OK) {
LOG(ERROR) << " Image reader creation failed on device model : "
<< base::android::BuildInfo::GetInstance()->model()
<< ". maxImages used is : " << max_images_;
base::debug::DumpWithoutCrashing();
if (return_code == AMEDIA_ERROR_INVALID_PARAMETER) {
LOG(ERROR) << "Either reader is null, or one or more of width, height, "
"format, maxImages arguments is not supported";
} else {
LOG(ERROR) << "unknown error";
}
return;
}
DCHECK(reader);
image_reader_ = reader;
// Create a new Image Listner.
listener_ = std::make_unique<AImageReader_ImageListener>();
// Passing |this| is safe here since we stop listening to new images in the
// destructor and set the ImageListener to null.
listener_->context = reinterpret_cast<void*>(this);
listener_->onImageAvailable = &ImageReaderGLOwner::OnFrameAvailable;
// Set the onImageAvailable listener of this image reader.
if (AImageReader_setImageListener(image_reader_, listener_.get()) !=
AMEDIA_OK) {
LOG(ERROR) << " Failed to register AImageReader listener";
return;
}
}
ImageReaderGLOwner::~ImageReaderGLOwner() {
DCHECK_CALLED_ON_VALID_THREAD(gpu_main_thread_checker_);
// Call ReleaseResources() if it hasn't already. This will do nothing if the
// texture and other resources has already been destroyed due to context loss.
ReleaseResources();
DCHECK_EQ(image_refs_.size(), 0u);
}
void ImageReaderGLOwner::ReleaseResources() {
DCHECK_CALLED_ON_VALID_THREAD(gpu_main_thread_checker_);
base::AutoLock auto_lock(lock_);
// Either TextureOwner is being destroyed or the TextureOwner's shared context
// is lost. Cleanup is it hasn't already.
if (image_reader_) {
// Now we can stop listening to new images.
AImageReader_setImageListener(image_reader_, nullptr);
// Delete all images before closing the associated image reader.
for (auto& image_ref : image_refs_)
AImage_delete(image_ref.first);
// Delete the image reader.
AImageReader_delete(image_reader_);
image_reader_ = nullptr;
// Clean up the ImageRefs which should now be a no-op since there is no
// valid |image_reader_|.
image_refs_.clear();
current_image_ref_.reset();
total_estimated_size_in_bytes_ = 0;
}
}
void ImageReaderGLOwner::SetFrameAvailableCallback(
const base::RepeatingClosure& frame_available_cb) {
DCHECK(!frame_available_cb_);
frame_available_cb_ = std::move(frame_available_cb);
}
gl::ScopedJavaSurface ImageReaderGLOwner::CreateJavaSurface() const {
base::AutoLock auto_lock(lock_);
// If we've already lost the texture, then do nothing.
if (!image_reader_) {
DLOG(ERROR) << "Already lost texture / image reader";
return nullptr;
}
// Get the android native window from the image reader.
ANativeWindow* window = nullptr;
if (AImageReader_getWindow(image_reader_, &window) != AMEDIA_OK) {
DLOG(ERROR) << "unable to get a window from image reader.";
return nullptr;
}
// Get the java surface object from the Android native window.
JNIEnv* env = base::android::AttachCurrentThread();
auto j_surface = base::android::ScopedJavaLocalRef<jobject>::Adopt(
env, ANativeWindow_toSurface(env, window));
DCHECK(j_surface);
// Get the scoped java surface that will call release() on destruction.
return gl::ScopedJavaSurface(j_surface, /*auto_release=*/true);
}
bool ImageReaderGLOwner::UpdateTexImage(bool discard) {
base::AutoLock auto_lock(lock_);
// If we've lost the texture, then do nothing.
if (!texture())
return false;
DCHECK(image_reader_);
// Acquire the latest image asynchronously. We must release the current image
// before acquiring a new one if the ImageReader was initialized with one
// outstanding image at max.
if (max_images_ == 1)
current_image_ref_.reset();
AImage* image = nullptr;
int acquire_fence_fd = -1;
media_status_t return_code = AMEDIA_OK;
if (max_images_ - image_refs_.size() < 2) {
// acquireNextImageAsync is required here since as per the spec calling
// AImageReader_acquireLatestImage with less than two images of margin, that
// is (maxImages - currentAcquiredImages < 2) will not discard as expected.
// We always have currentAcquiredImages as 1 since we delete a previous
// image only after acquiring a new image.
return_code = AImageReader_acquireNextImageAsync(image_reader_, &image,
&acquire_fence_fd);
} else {
return_code = AImageReader_acquireLatestImageAsync(image_reader_, &image,
&acquire_fence_fd);
}
base::UmaHistogramSparse("Media.AImageReaderGLOwner.AcquireImageResult",
return_code);
UMA_HISTOGRAM_ENUMERATION("Media.AImageReaderGLOwner.CodecType",
type_for_metrics_);
// TODO(http://crbug.com/846050).
// Need to add some better error handling if below error occurs. Currently we
// just return if error occurs.
switch (return_code) {
case AMEDIA_ERROR_INVALID_PARAMETER:
LOG(ERROR) << "AImageReader: Invalid parameter";
return false;
case AMEDIA_IMGREADER_MAX_IMAGES_ACQUIRED:
LOG(ERROR)
<< "number of concurrently acquired images has reached the limit";
return false;
case AMEDIA_IMGREADER_NO_BUFFER_AVAILABLE:
LOG(ERROR) << "no buffers currently available in the reader queue";
return false;
case AMEDIA_ERROR_UNKNOWN:
LOG(ERROR) << "method fails for some other reasons";
return false;
case AMEDIA_OK:
// Method call succeeded.
break;
default:
LOG(ERROR) << "AImageReader: Unknown error: " << return_code;
// No other error code should be returned.
NOTREACHED();
}
base::ScopedFD scoped_acquire_fence_fd(acquire_fence_fd);
// If there is no new image simply return. At this point previous image will
// still be bound to the texture.
if (!image) {
LOG(ERROR) << "AImageReader: image is nullptr: " << return_code;
return false;
}
UMA_HISTOGRAM_BOOLEAN("Media.AImageReaderGLOwner.HasFence",
scoped_acquire_fence_fd.is_valid());
if (discard && current_image_ref_ &&
base::FeatureList::IsEnabled(kDiscardDroppedEarlyRenderedFrames)) {
if (scoped_acquire_fence_fd.is_valid()) {
AImage_deleteAsync(image, scoped_acquire_fence_fd.release());
} else {
AImage_delete(image);
}
return true;
}
// Make the newly acquired image as current image.
current_image_ref_.emplace(this, image, std::move(scoped_acquire_fence_fd));
return true;
}
std::unique_ptr<base::android::ScopedHardwareBufferFenceSync>
ImageReaderGLOwner::GetAHardwareBuffer() {
base::AutoLock auto_lock(lock_);
if (!current_image_ref_)
return nullptr;
AHardwareBuffer* buffer = nullptr;
auto error = AImage_getHardwareBuffer(current_image_ref_->image(), &buffer);
if (!buffer) {
LOG(ERROR) << "AImage_getHardwareBuffer returned nullptr: " << error;
return nullptr;
}
// TODO(crbug.com/40749597): We suspect that buffer is already freed here and
// it causes crash later. Trying to crash earlier.
base::AndroidHardwareBufferCompat::GetInstance().Acquire(buffer);
base::AndroidHardwareBufferCompat::GetInstance().Release(buffer);
return std::make_unique<ScopedHardwareBufferImpl>(
this, current_image_ref_->image(),
base::android::ScopedHardwareBufferHandle::Create(buffer),
current_image_ref_->GetReadyFence());
}
gfx::Rect ImageReaderGLOwner::GetCropRectLocked() {
lock_.AssertAcquired();
if (!current_image_ref_)
return gfx::Rect();
// Note that to query the crop rectangle, we don't need to wait for the
// AImage to be ready by checking the associated image ready fence.
AImageCropRect crop_rect;
media_status_t return_code =
AImage_getCropRect(current_image_ref_->image(), &crop_rect);
if (return_code != AMEDIA_OK) {
DLOG(ERROR) << "Error querying crop rectangle from the image : "
<< return_code;
return gfx::Rect();
}
DCHECK_GE(crop_rect.right, crop_rect.left);
DCHECK_GE(crop_rect.bottom, crop_rect.top);
return gfx::Rect(crop_rect.left, crop_rect.top,
crop_rect.right - crop_rect.left,
crop_rect.bottom - crop_rect.top);
}
void ImageReaderGLOwner::RegisterRefOnImageLocked(AImage* image) {
lock_.AssertAcquired();
DCHECK(image_reader_);
auto& ref = image_refs_[image];
// Add a ref that the caller will release.
if (ref.count++ == 0) {
if (auto size = GetImageSize(image)) {
ref.size = size.value();
// We don't know the exact format of the image so we use NV12 as
// approximation as the most popular format.
constexpr auto format = viz::MultiPlaneFormat::kNV12;
ref.estimated_size_in_bytes = format.EstimatedSizeInBytes(ref.size);
total_estimated_size_in_bytes_ += ref.estimated_size_in_bytes;
}
}
}
void ImageReaderGLOwner::ReleaseRefOnImage(AImage* image,
base::ScopedFD fence_fd) {
base::AutoLock auto_lock(lock_);
ReleaseRefOnImageLocked(image, std::move(fence_fd));
}
void ImageReaderGLOwner::ReleaseRefOnImageLocked(AImage* image,
base::ScopedFD fence_fd) {
lock_.AssertAcquired();
// During cleanup on losing the texture, all images are synchronously released
// and the |image_reader_| is destroyed.
if (!image_reader_)
return;
// Ensure that DrDc lock is held when |buffer_available_cb| can be triggered
// because we do not want any other thread to steal the free buffer slot which
// is meant to be used by |buffer_available_cb| and hence resulting in wrong
// FrameInfo for all future frames.
AssertAcquiredDrDcLock();
auto it = image_refs_.find(image);
CHECK(it != image_refs_.end());
auto& image_ref = it->second;
DCHECK_GT(image_ref.count, 0u);
image_ref.count--;
image_ref.release_fence_fd =
gl::MergeFDs(std::move(image_ref.release_fence_fd), std::move(fence_fd));
if (image_ref.count > 0)
return;
if (image_ref.release_fence_fd.is_valid()) {
AImage_deleteAsync(image, std::move(image_ref.release_fence_fd.release()));
} else {
AImage_delete(image);
}
total_estimated_size_in_bytes_ -= it->second.estimated_size_in_bytes;
image_refs_.erase(it);
DCHECK_GT(max_images_, static_cast<int32_t>(image_refs_.size()));
auto buffer_available_cb = std::move(buffer_available_cb_);
// |buffer_available_cb| will try to acquire lock again via
// UpdatetexImage(), hence we need to unlock here. Note that when
// |max_images_| is 1, this callback will always be empty here since it will
// be run immediately in RunWhenBufferIsAvailable(). Hence resetting
// |current_image_ref_| in UpdateTexImage() can not trigger this callback.
// Otherwise triggering this callback from UpdateTexImage() on
// |current_image_ref_| reset would cause callback and hence FrameInfoHelper
// to run and eventually call UpdateTexImage() from there which could have
// been filmsy.
if (buffer_available_cb) {
base::AutoUnlock auto_unlock(lock_);
DCHECK_GT(max_images_, 1);
std::move(buffer_available_cb).Run();
}
}
void ImageReaderGLOwner::ReleaseBackBuffers() {
DCHECK_CALLED_ON_VALID_THREAD(gpu_main_thread_checker_);
// ReleaseBackBuffers() call is not required with image reader.
}
gl::GLContext* ImageReaderGLOwner::GetContext() const {
DCHECK_CALLED_ON_VALID_THREAD(gpu_main_thread_checker_);
return context_.get();
}
gl::GLSurface* ImageReaderGLOwner::GetSurface() const {
DCHECK_CALLED_ON_VALID_THREAD(gpu_main_thread_checker_);
return surface_.get();
}
// This callback function will be called when there is a new image available
// for in the image reader's queue.
void ImageReaderGLOwner::OnFrameAvailable(void* context, AImageReader* reader) {
ImageReaderGLOwner* image_reader_ptr =
reinterpret_cast<ImageReaderGLOwner*>(context);
// It is safe to run this callback on any thread.
image_reader_ptr->frame_available_cb_.Run();
}
void ImageReaderGLOwner::RunWhenBufferIsAvailable(base::OnceClosure callback) {
DCHECK_CALLED_ON_VALID_THREAD(gpu_main_thread_checker_);
int image_refs_size = 0;
{
base::AutoLock auto_lock(lock_);
// Note that we handle only one simultaneous request, this is not issue
// because FrameInfoHelper maintain request queue and has only single
// outstanding request on GPU thread.
DCHECK(!buffer_available_cb_);
image_refs_size = static_cast<int>(image_refs_.size());
}
// If `max_images` == 1 we will drop it before acquiring new buffer. Note
// that this must never happen with SurfaceControl and the
// ImageReaderGLOwner is the sole owner of the images.
if (max_images_ == 1 || image_refs_size < max_images_) {
// This callback is run from here as well as from ReleaseRefOnImage() where
// we remove one image from image reader queue before callback is run.
// Once the |lock_| is dropped in this method here, another thread can
// UpdateTexImage() before callback is run and hence cause the image reader
// queue to become full. In that case callback will not be able to render
// and acquire updated image and hence will use FrameInfo of the previous
// image which will result in wrong coded size for all future frames. To
// avoid this, no other thread should try to UpdateTexImage() when this
// callback is run. Hence drdc_lock should be held from all the places from
// where the callback could be run which is either OnGpu::GetFrameInfo() or
// ImageReaderGLOwner::ReleaseRefOnImageLocked() and
// OnGpu::GetFrameInfoImpl() should assume that the drdc_lock is always
// held.
std::move(callback).Run();
} else {
base::AutoLock auto_lock(lock_);
buffer_available_cb_ = std::move(callback);
}
}
bool ImageReaderGLOwner::GetCodedSizeAndVisibleRect(
gfx::Size rotated_visible_size,
gfx::Size* coded_size,
gfx::Rect* visible_rect) {
base::AutoLock auto_lock(lock_);
DCHECK(visible_rect);
DCHECK(coded_size);
AHardwareBuffer* buffer = nullptr;
if (current_image_ref_) {
AImage_getHardwareBuffer(current_image_ref_->image(), &buffer);
if (!buffer) {
DLOG(ERROR) << "Unable to get an AHardwareBuffer from the image";
}
}
if (!buffer) {
*coded_size = gfx::Size();
*visible_rect = gfx::Rect();
return false;
}
// Get the buffer descriptor. Note that for querying the buffer descriptor, we
// do not need to wait on the AHB to be ready.
AHardwareBuffer_Desc desc;
base::AndroidHardwareBufferCompat::GetInstance().Describe(buffer, &desc);
*visible_rect = GetCropRectLocked();
*coded_size = gfx::Size(desc.width, desc.height);
return true;
}
bool ImageReaderGLOwner::OnMemoryDump(
const base::trace_event::MemoryDumpArgs& args,
base::trace_event::ProcessMemoryDump* pmd) {
if (args.level_of_detail ==
base::trace_event::MemoryDumpLevelOfDetail::kBackground) {
auto dump_name = base::StringPrintf(kMemoryDumpPrefix, tracing_id());
base::trace_event::MemoryAllocatorDump* dump =
pmd->CreateAllocatorDump(dump_name);
dump->AddScalar(base::trace_event::MemoryAllocatorDump::kNameSize,
base::trace_event::MemoryAllocatorDump::kUnitsBytes,
total_estimated_size_in_bytes_);
// Early out, no need for more detail in a BACKGROUND dump.
return true;
}
int i = 0;
base::AutoLock auto_lock(lock_);
for (const auto& image : image_refs_) {
std::string dump_name = base::StringPrintf(
"gpu/media_texture_owner_0x%x/image_%d", tracing_id(), i++);
// If we fail to get AImage size for any reason, we still report the image
// as a empty size, so it can be diagnosed in necessary.
base::trace_event::MemoryAllocatorDump* dump =
pmd->CreateAllocatorDump(dump_name);
dump->AddScalar(base::trace_event::MemoryAllocatorDump::kNameSize,
base::trace_event::MemoryAllocatorDump::kUnitsBytes,
image.second.estimated_size_in_bytes);
dump->AddString("dimensions", "", image.second.size.ToString());
}
return true;
}
ImageReaderGLOwner::ImageRef::ImageRef() = default;
ImageReaderGLOwner::ImageRef::~ImageRef() = default;
ImageReaderGLOwner::ImageRef::ImageRef(ImageRef&& other) = default;
ImageReaderGLOwner::ImageRef& ImageReaderGLOwner::ImageRef::operator=(
ImageRef&& other) = default;
ImageReaderGLOwner::ScopedCurrentImageRef::ScopedCurrentImageRef(
ImageReaderGLOwner* texture_owner,
AImage* image,
base::ScopedFD ready_fence)
: texture_owner_(texture_owner),
image_(image),
ready_fence_(std::move(ready_fence)) {
DCHECK(texture_owner_);
texture_owner_->lock_.AssertAcquired();
DCHECK(image_);
texture_owner_->RegisterRefOnImageLocked(image_);
}
ImageReaderGLOwner::ScopedCurrentImageRef::~ScopedCurrentImageRef() {
texture_owner_->lock_.AssertAcquired();
texture_owner_->ReleaseRefOnImageLocked(image_, std::move(ready_fence_));
}
base::ScopedFD ImageReaderGLOwner::ScopedCurrentImageRef::GetReadyFence()
const {
return base::ScopedFD(HANDLE_EINTR(dup(ready_fence_.get())));
}
} // namespace gpu
|