Chromium Code Reviews
chromiumcodereview-hr@appspot.gserviceaccount.com (chromiumcodereview-hr) | Please choose your nickname with Settings | Help | Chromium Project | Gerrit Changes | Sign out
(1297)

Unified Diff: content/gpu/gpu_main.cc

Issue 2322123002: gpu: Introduce GpuInit. (Closed)
Patch Set: . Created 4 years, 3 months ago
Use n/p to move between diff chunks; N/P to move between comments. Draft comments are only viewable by you.
Jump to:
View side-by-side diff with in-line comments
Download patch
« no previous file with comments | « content/gpu/gpu_child_thread.h ('k') | content/public/common/content_switches.h » ('j') | no next file with comments »
Expand Comments ('e') | Collapse Comments ('c') | Show Comments Hide Comments ('s')
Index: content/gpu/gpu_main.cc
diff --git a/content/gpu/gpu_main.cc b/content/gpu/gpu_main.cc
index 6e1c7326ec3a5858cf1c6b62da7a8d12e8ec14ed..e76234bf3decc27388333f493c06000123f038d6 100644
--- a/content/gpu/gpu_main.cc
+++ b/content/gpu/gpu_main.cc
@@ -36,6 +36,7 @@
#include "gpu/config/gpu_util.h"
#include "gpu/ipc/common/gpu_memory_buffer_support.h"
#include "gpu/ipc/service/gpu_config.h"
+#include "gpu/ipc/service/gpu_init.h"
#include "gpu/ipc/service/gpu_memory_buffer_factory.h"
#include "gpu/ipc/service/gpu_watchdog_thread.h"
#include "ui/events/platform/platform_event_source.h"
@@ -91,19 +92,8 @@ namespace content {
namespace {
-void GetGpuInfoFromCommandLine(gpu::GPUInfo& gpu_info,
- const base::CommandLine& command_line);
-void WarmUpSandbox();
-
-#if !defined(OS_MACOSX)
-bool CollectGraphicsInfo(gpu::GPUInfo& gpu_info);
-#endif
-
#if defined(OS_LINUX)
-#if !defined(OS_CHROMEOS)
-bool CanAccessNvidiaDeviceFile();
-#endif
-bool StartSandboxLinux(const gpu::GPUInfo&, gpu::GpuWatchdogThread*);
+bool StartSandboxLinux(gpu::GpuWatchdogThread*);
#elif defined(OS_WIN)
bool StartSandboxWindows(const sandbox::SandboxInterfaceInfo*);
#endif
@@ -122,6 +112,62 @@ bool GpuProcessLogMessageHandler(int severity,
return false;
}
+class ContentSandboxHelper : public gpu::GpuSandboxHelper {
+ public:
+ ContentSandboxHelper() {}
+ ~ContentSandboxHelper() override {}
+
+#if defined(OS_WIN)
+ void set_sandbox_info(const sandbox::SandboxInterfaceInfo* info) {
+ sandbox_info_ = info;
+ }
+#endif
+
+#if defined(OS_LINUX)
+ void set_gpu_init(gpu::GpuInit* gpu_init) { gpu_init_ = gpu_init; }
+#endif
+
+ private:
+ // SandboxHelper:
+ void PreSandboxStartup() override {
+ // Warm up resources that don't need access to GPUInfo.
+ {
+ TRACE_EVENT0("gpu", "Warm up rand");
+ // Warm up the random subsystem, which needs to be done pre-sandbox on all
+ // platforms.
+ (void)base::RandUint64();
+ }
+
+#if defined(OS_CHROMEOS) && defined(ARCH_CPU_X86_FAMILY)
+ media::VaapiWrapper::PreSandboxInitialization();
+#endif
+#if defined(OS_WIN)
+ media::DXVAVideoDecodeAccelerator::PreSandboxInitialization();
+ media::MediaFoundationVideoEncodeAccelerator::PreSandboxInitialization();
+#endif
+ }
+
+ bool EnsureSandboxInitialized() override {
+#if defined(OS_LINUX)
+ return StartSandboxLinux(gpu_init_->watchdog_thread());
+#elif defined(OS_WIN)
+ return StartSandboxWindows(sandbox_info_);
+#elif defined(OS_MACOSX)
+ return Sandbox::SandboxIsCurrentlyActive();
+#else
+ return false;
+#endif
+ }
+
+#if defined(OS_WIN)
+ const sandbox::SandboxInterfaceInfo* sandbox_info_ = nullptr;
+#elif defined(OS_LINUX)
+ gpu::GpuInit* gpu_init_ = nullptr;
+#endif
+
+ DISALLOW_COPY_AND_ASSIGN(ContentSandboxHelper);
+};
+
} // namespace anonymous
// Main function for starting the Gpu process.
@@ -159,24 +205,6 @@ int GpuMain(const MainFunctionParams& parameters) {
logging::SetLogMessageHandler(GpuProcessLogMessageHandler);
- if (command_line.HasSwitch(switches::kSupportsDualGpus)) {
- std::set<int> workarounds;
- gpu::GpuDriverBugList::AppendWorkaroundsFromCommandLine(&workarounds,
- command_line);
- gpu::InitializeDualGpusIfSupported(workarounds);
- }
-
- // Initialization of the OpenGL bindings may fail, in which case we
- // will need to tear down this process. However, we can not do so
- // safely until the IPC channel is set up, because the detection of
- // early return of a child process is implemented using an IPC
- // channel error. If the IPC channel is not fully set up between the
- // browser and GPU process, and the GPU process crashes or exits
- // early, the browser process will never detect it. For this reason
- // we defer tearing down the GPU process until receiving the
- // GpuMsg_Initialize message from the browser.
- bool dead_on_arrival = false;
-
#if defined(OS_WIN)
// Use a UI message loop because ANGLE and the desktop GL platform can
// create child windows to render to.
@@ -201,158 +229,41 @@ int GpuMain(const MainFunctionParams& parameters) {
base::PlatformThread::SetName("CrGpuMain");
- // In addition to disabling the watchdog if the command line switch is
- // present, disable the watchdog on valgrind because the code is expected
- // to run slowly in that case.
- bool enable_watchdog =
- !command_line.HasSwitch(switches::kDisableGpuWatchdog) &&
- !RunningOnValgrind();
-
- // Disable the watchdog in debug builds because they tend to only be run by
- // developers who will not appreciate the watchdog killing the GPU process.
-#ifndef NDEBUG
- enable_watchdog = false;
-#endif
-
- bool delayed_watchdog_enable = false;
-
-#if defined(OS_CHROMEOS)
- // Don't start watchdog immediately, to allow developers to switch to VT2 on
- // startup.
- delayed_watchdog_enable = true;
-#endif
-
- scoped_refptr<gpu::GpuWatchdogThread> watchdog_thread;
-
- // Start the GPU watchdog only after anything that is expected to be time
- // consuming has completed, otherwise the process is liable to be aborted.
- if (enable_watchdog && !delayed_watchdog_enable)
- watchdog_thread = gpu::GpuWatchdogThread::Create();
-
// Initializes StatisticsRecorder which tracks UMA histograms.
base::StatisticsRecorder::Initialize();
- gpu::GPUInfo gpu_info;
- // Get vendor_id, device_id, driver_version from browser process through
- // commandline switches.
- GetGpuInfoFromCommandLine(gpu_info, command_line);
- gpu_info.in_process_gpu = false;
-
-#if defined(OS_CHROMEOS) && defined(ARCH_CPU_X86_FAMILY)
- media::VaapiWrapper::PreSandboxInitialization();
-#endif
-
#if defined(OS_ANDROID) || defined(OS_CHROMEOS)
// Set thread priority before sandbox initialization.
base::PlatformThread::SetCurrentThreadPriority(base::ThreadPriority::DISPLAY);
#endif
- // Warm up resources that don't need access to GPUInfo.
- WarmUpSandbox();
-
-#if defined(OS_LINUX)
- bool initialized_sandbox = false;
- // On Chrome OS ARM Mali, GPU driver userspace creates threads when
- // initializing a GL context, so start the sandbox early.
- if (command_line.HasSwitch(switches::kGpuSandboxStartEarly)) {
- gpu_info.sandboxed = StartSandboxLinux(gpu_info, watchdog_thread.get());
- initialized_sandbox = true;
- }
-#endif // defined(OS_LINUX)
-
- base::TimeTicks before_initialize_one_off = base::TimeTicks::Now();
-
- // Determine if we need to initialize GL here or it has already been done.
- bool gl_already_initialized = false;
-#if defined(OS_MACOSX)
- if (!command_line.HasSwitch(switches::kNoSandbox)) {
- // On Mac, if the sandbox is enabled, then gl::init::InitializeGLOneOff()
- // is called from the sandbox warmup code before getting here.
- gl_already_initialized = true;
- }
+ gpu::GpuInit gpu_init;
+ ContentSandboxHelper sandbox_helper;
+#if defined(OS_WIN)
+ sandbox_helper.set_sandbox_info(parameters.sandbox_info);
+#elif defined(OS_LINUX)
+ sandbox_helper.set_gpu_init(&gpu_init);
#endif
- if (command_line.HasSwitch(switches::kInProcessGPU)) {
- // With in-process GPU, gl::init::InitializeGLOneOff() is called from
- // GpuChildThread before getting here.
- gl_already_initialized = true;
- }
- // Load and initialize the GL implementation and locate the GL entry points.
- bool gl_initialized =
- gl_already_initialized
- ? gl::GetGLImplementation() != gl::kGLImplementationNone
- : gl::init::InitializeGLOneOff();
- if (gl_initialized) {
- // We need to collect GL strings (VENDOR, RENDERER) for blacklisting
- // purposes. However, on Mac we don't actually use them. As documented in
- // crbug.com/222934, due to some driver issues, glGetString could take
- // multiple seconds to finish, which in turn cause the GPU process to
- // crash.
- // By skipping the following code on Mac, we don't really lose anything,
- // because the basic GPU information is passed down from browser process
- // and we already registered them through SetGpuInfo() above.
- base::TimeTicks before_collect_context_graphics_info =
- base::TimeTicks::Now();
-#if !defined(OS_MACOSX)
- if (!CollectGraphicsInfo(gpu_info))
- dead_on_arrival = true;
-
- // Recompute gpu driver bug workarounds.
- // This is necessary on systems where vendor_id/device_id aren't available
- // (Chrome OS, Android) or where workarounds may be dependent on GL_VENDOR
- // and GL_RENDERER strings which are lazily computed (Linux).
- if (!command_line.HasSwitch(switches::kDisableGpuDriverBugWorkarounds)) {
- // TODO: this can not affect disabled extensions, since they're already
- // initialized in the bindings. This should be moved before bindings
- // initialization. However, populating GPUInfo fully works only on
- // Android. Other platforms would need the bindings to query GL strings.
- gpu::ApplyGpuDriverBugWorkarounds(
- gpu_info, const_cast<base::CommandLine*>(&command_line));
- }
-#endif // !defined(OS_MACOSX)
-
-#if defined(OS_LINUX) && !defined(OS_CHROMEOS)
- if (gpu_info.gpu.vendor_id == 0x10de && // NVIDIA
- gpu_info.driver_vendor == "NVIDIA" && !CanAccessNvidiaDeviceFile())
- dead_on_arrival = true;
-#endif // defined(OS_LINUX) && !defined(OS_CHROMEOS)
-
- base::TimeDelta collect_context_time =
- base::TimeTicks::Now() - before_collect_context_graphics_info;
- UMA_HISTOGRAM_TIMES("GPU.CollectContextGraphicsInfo", collect_context_time);
- } else { // gl_initialized
- VLOG(1) << "gl::init::InitializeGLOneOff failed";
- dead_on_arrival = true;
- }
+ gpu_init.set_sandbox_helper(&sandbox_helper);
- base::TimeDelta initialize_one_off_time =
- base::TimeTicks::Now() - before_initialize_one_off;
- UMA_HISTOGRAM_MEDIUM_TIMES("GPU.InitializeOneOffMediumTime",
- initialize_one_off_time);
- if (enable_watchdog && delayed_watchdog_enable)
- watchdog_thread = gpu::GpuWatchdogThread::Create();
-
- // OSMesa is expected to run very slowly, so disable the watchdog in that
- // case.
- if (enable_watchdog &&
- gl::GetGLImplementation() == gl::kGLImplementationOSMesaGL) {
- watchdog_thread->Stop();
- watchdog_thread = NULL;
- }
-
-#if defined(OS_LINUX)
- if (!initialized_sandbox)
- gpu_info.sandboxed = StartSandboxLinux(gpu_info, watchdog_thread.get());
-#elif defined(OS_WIN)
- gpu_info.sandboxed = StartSandboxWindows(parameters.sandbox_info);
-#elif defined(OS_MACOSX)
- gpu_info.sandboxed = Sandbox::SandboxIsCurrentlyActive();
-#endif
+ // Gpu initialization may fail for various reasons, in which case we will need
+ // to tear down this process. However, we can not do so safely until the IPC
+ // channel is set up, because the detection of early return of a child process
+ // is implemented using an IPC channel error. If the IPC channel is not fully
+ // set up between the browser and GPU process, and the GPU process crashes or
+ // exits early, the browser process will never detect it. For this reason we
+ // defer tearing down the GPU process until receiving the GpuMsg_Initialize
+ // message from the browser.
+ const bool init_success = gpu_init.InitializeAndStartSandbox(command_line);
+ const bool dead_on_arrival = !init_success;
logging::SetLogMessageHandler(NULL);
+ GetContentClient()->SetGpuInfo(gpu_init.gpu_info());
std::unique_ptr<gpu::GpuMemoryBufferFactory> gpu_memory_buffer_factory;
- if (gpu::GetNativeGpuMemoryBufferType() != gfx::EMPTY_BUFFER)
+ if (init_success &&
+ gpu::GetNativeGpuMemoryBufferType() != gfx::EMPTY_BUFFER)
gpu_memory_buffer_factory = gpu::GpuMemoryBufferFactory::CreateNativeType();
base::ThreadPriority io_thread_priority = base::ThreadPriority::NORMAL;
@@ -361,10 +272,9 @@ int GpuMain(const MainFunctionParams& parameters) {
#endif
GpuProcess gpu_process(io_thread_priority);
-
GpuChildThread* child_thread = new GpuChildThread(
- watchdog_thread.get(), dead_on_arrival, gpu_info, deferred_messages.Get(),
- gpu_memory_buffer_factory.get());
+ gpu_init.watchdog_thread(), dead_on_arrival, gpu_init.gpu_info(),
+ deferred_messages.Get(), gpu_memory_buffer_factory.get());
while (!deferred_messages.Get().empty())
deferred_messages.Get().pop();
@@ -372,8 +282,8 @@ int GpuMain(const MainFunctionParams& parameters) {
gpu_process.set_main_thread(child_thread);
- if (watchdog_thread.get())
- watchdog_thread->AddPowerObserver();
+ if (gpu_init.watchdog_thread())
+ gpu_init.watchdog_thread()->AddPowerObserver();
#if defined(OS_ANDROID)
base::trace_event::MemoryDumpManager::GetInstance()->RegisterDumpProvider(
@@ -393,111 +303,8 @@ int GpuMain(const MainFunctionParams& parameters) {
namespace {
-void GetGpuInfoFromCommandLine(gpu::GPUInfo& gpu_info,
- const base::CommandLine& command_line) {
- DCHECK(command_line.HasSwitch(switches::kGpuVendorID) &&
- command_line.HasSwitch(switches::kGpuDeviceID) &&
- command_line.HasSwitch(switches::kGpuDriverVersion));
- bool success = base::HexStringToUInt(
- command_line.GetSwitchValueASCII(switches::kGpuVendorID),
- &gpu_info.gpu.vendor_id);
- DCHECK(success);
- success = base::HexStringToUInt(
- command_line.GetSwitchValueASCII(switches::kGpuDeviceID),
- &gpu_info.gpu.device_id);
- DCHECK(success);
- gpu_info.driver_vendor =
- command_line.GetSwitchValueASCII(switches::kGpuDriverVendor);
- gpu_info.driver_version =
- command_line.GetSwitchValueASCII(switches::kGpuDriverVersion);
- gpu_info.driver_date =
- command_line.GetSwitchValueASCII(switches::kGpuDriverDate);
- gpu::ParseSecondaryGpuDevicesFromCommandLine(command_line, &gpu_info);
-
- // Set active gpu device.
- if (command_line.HasSwitch(switches::kGpuActiveVendorID) &&
- command_line.HasSwitch(switches::kGpuActiveDeviceID)) {
- uint32_t active_vendor_id = 0;
- uint32_t active_device_id = 0;
- success = base::HexStringToUInt(
- command_line.GetSwitchValueASCII(switches::kGpuActiveVendorID),
- &active_vendor_id);
- DCHECK(success);
- success = base::HexStringToUInt(
- command_line.GetSwitchValueASCII(switches::kGpuActiveDeviceID),
- &active_device_id);
- DCHECK(success);
- if (gpu_info.gpu.vendor_id == active_vendor_id &&
- gpu_info.gpu.device_id == active_device_id) {
- gpu_info.gpu.active = true;
- } else {
- for (size_t i = 0; i < gpu_info.secondary_gpus.size(); ++i) {
- if (gpu_info.secondary_gpus[i].vendor_id == active_vendor_id &&
- gpu_info.secondary_gpus[i].device_id == active_device_id) {
- gpu_info.secondary_gpus[i].active = true;
- break;
- }
- }
- }
- }
-
- GetContentClient()->SetGpuInfo(gpu_info);
-}
-
-void WarmUpSandbox() {
- {
- TRACE_EVENT0("gpu", "Warm up rand");
- // Warm up the random subsystem, which needs to be done pre-sandbox on all
- // platforms.
- (void) base::RandUint64();
- }
-
-#if defined(OS_WIN)
- media::DXVAVideoDecodeAccelerator::PreSandboxInitialization();
- media::MediaFoundationVideoEncodeAccelerator::PreSandboxInitialization();
-#endif
-}
-
-#if !defined(OS_MACOSX)
-bool CollectGraphicsInfo(gpu::GPUInfo& gpu_info) {
- TRACE_EVENT0("gpu,startup", "Collect Graphics Info");
-
- bool res = true;
- gpu::CollectInfoResult result = gpu::CollectContextGraphicsInfo(&gpu_info);
- switch (result) {
- case gpu::kCollectInfoFatalFailure:
- LOG(ERROR) << "gpu::CollectGraphicsInfo failed (fatal).";
- res = false;
- break;
- case gpu::kCollectInfoNonFatalFailure:
- DVLOG(1) << "gpu::CollectGraphicsInfo failed (non-fatal).";
- break;
- case gpu::kCollectInfoNone:
- NOTREACHED();
- break;
- case gpu::kCollectInfoSuccess:
- break;
- }
- GetContentClient()->SetGpuInfo(gpu_info);
- return res;
-}
-#endif
-
#if defined(OS_LINUX)
-#if !defined(OS_CHROMEOS)
-bool CanAccessNvidiaDeviceFile() {
- bool res = true;
- base::ThreadRestrictions::AssertIOAllowed();
- if (access("/dev/nvidiactl", R_OK) != 0) {
- DVLOG(1) << "NVIDIA device file /dev/nvidiactl access denied";
- res = false;
- }
- return res;
-}
-#endif
-
-bool StartSandboxLinux(const gpu::GPUInfo& gpu_info,
- gpu::GpuWatchdogThread* watchdog_thread) {
+bool StartSandboxLinux(gpu::GpuWatchdogThread* watchdog_thread) {
TRACE_EVENT0("gpu,startup", "Initialize sandbox");
bool res = false;
« no previous file with comments | « content/gpu/gpu_child_thread.h ('k') | content/public/common/content_switches.h » ('j') | no next file with comments »

Powered by Google App Engine
This is Rietveld 408576698