Skip to content

A
angle
Switch branch/tag
  1. 15 Oct, 2010 4 commits
  3. 14 Oct, 2010 6 commits
  5. 13 Oct, 2010 1 commit
  7. 07 Oct, 2010 2 commits
  9. 06 Oct, 2010 1 commit
  11. 05 Oct, 2010 1 commit
    • Fixed NULL dereference in glClear. · b2bdd06c
      apatrick@chromium.org authored 
      6 crashes on Windows XP
1 crash on Windows 7

This was Canary 7.0.530.0 using ANGLE r429.

0x011f523b 	[libGLESv2.dll 	- context.cpp:2412] 	gl::Context::clear(unsigned int)
0x011e5f41 	[libGLESv2.dll 	- libglesv2.cpp:611] 	glClear
0x020c400d 	[chrome.dll 	- gl_context.cc:33] 	gfx::GLContext::InitializeCommon()
0x020bb757 	[chrome.dll 	- gl_context_egl.cc:138] 	gfx::NativeViewEGLContext::Initialize()
0x020b8946 	[chrome.dll 	- gl_context_win.cc:502] 	gfx::GLContext::CreateViewGLContext(HWND__ *,bool)
0x01c4f6f8 	[chrome.dll 	- gpu_processor_win.cc:35] 	gpu::GPUProcessor::Initialize(HWND__ *,gfx::Size const &,gpu::GPUProcessor *,unsigned int)
0x01d6e669 	[chrome.dll 	- gpu_command_buffer_stub.cc:88] 	GpuCommandBufferStub::OnInitialize(int,void * *)

It's crashing upon creation of the GL context, possibly the first context. Device lost would possibly explain the XP crashes. The Windows 7 box might have run out of video memory.

I also checked another couple of NULL dereference crashes.


Review URL: http://codereview.appspot.com/2197046

git-svn-id: https://angleproject.googlecode.com/svn/trunk@444 736b8ea6-26fd-11df-bfd4-992fa37f6226
  13. 01 Oct, 2010 1 commit
  15. 28 Sep, 2010 1 commit
  17. 27 Sep, 2010 2 commits
  19. 24 Sep, 2010 1 commit
  21. 23 Sep, 2010 2 commits
  23. 22 Sep, 2010 4 commits
  25. 21 Sep, 2010 2 commits
  27. 20 Sep, 2010 3 commits
    • I hope this will fix a generateMipmaps crash when needRenderTarget fails. · 9398a6b2
      apatrick@chromium.org authored 
      This fixes this crash in 7.0.524.0 on a user's Windows XP service pack 3 box.

0x014946f3 	[libGLESv2.dll 	- texture.cpp:1382] 	gl::Texture2D::generateMipmaps()
0x0149786f 	[libGLESv2.dll 	- libglesv2.cpp:2114] 	glGenerateMipmap
0x01c53199 	[chrome.dll 	- gles2_cmd_decoder.cc:2437] 	gpu::gles2::GLES2DecoderImpl::DoGenerateMipmap(unsigned int)
0x01c52216 	[chrome.dll 	- gles2_cmd_decoder.cc:2214] 	gpu::gles2::GLES2DecoderImpl::DoCommand(unsigned int,unsigned int,void const *)
0x01c4fa78 	[chrome.dll 	- cmd_parser.cc:56] 	gpu::CommandParser::ProcessCommand()
0x01c4f8ba 	[chrome.dll 	- gpu_processor.cc:100] 	gpu::GPUProcessor::ProcessCommands()
0x01ecdcce 	[chrome.dll 	- callback.h:119] 	CallbackImpl<BrowsingDataRemover,void ( BrowsingDataRemover::*)(void),Tuple0>::RunWithParams(Tuple0 const &)
0x01c4f2d9 	[chrome.dll 	- callback.h:79] 	CallbackRunner<Tuple0>::Run()
0x01c4edce 	[chrome.dll 	- command_buffer_service.cc:80] 	gpu::CommandBufferService::Flush(int)
0x01d6b9b1 	[chrome.dll 	- gpu_command_buffer_stub.cc:129] 	GpuCommandBufferStub::OnFlush(int,gpu::CommandBuffer::State *)

This might be glGenerateMipmaps after a device lost.


Review URL: http://codereview.appspot.com/2227044

git-svn-id: https://angleproject.googlecode.com/svn/trunk@429 736b8ea6-26fd-11df-bfd4-992fa37f6226
    • Trying to fix bug crash when eglCreateWindowSurface (or any other code that… · e860f079
      apatrick@chromium.org authored 
      Trying to fix bug crash when eglCreateWindowSurface (or any other code that causes createDevice to be invoked) is called and CreateDevice reports DEVICELOST or NOTAVAILABLE.

This was a Windows XP service pack 3 box.

0x10002545 	[libegl.dll 	- display.cpp:340] 	egl::Display::createDevice()
0x10002912 	[libegl.dll 	- display.cpp:468] 	egl::Display::getDevice()
0x01d6a80f 	[chrome.dll 	- gpu_info_collector_win.cc:29] 	gpu_info_collector::CollectGraphicsInfo(GPUInfo *)
0x01d68eca 	[chrome.dll 	- gpu_thread.cc:104] 	GpuThread::OnEstablishChannel(int)

Previously createDevice only checked for out of memory and the ASSERT(SUCCEEDED(result)) has no effect in release builds.

I simulated the failure in a debugger and discovered a second place where a null dereference would occur in this case.

I don't think this crash is technically correct but it prevents crashes.
Review URL: http://codereview.appspot.com/2217043

git-svn-id: https://angleproject.googlecode.com/svn/trunk@428 736b8ea6-26fd-11df-bfd4-992fa37f6226
    • Texture2D::getRenderTarget does not assume that needsRenderTarget succeeded. · 4d5962cd
      apatrick@chromium.org authored 
      Specifically, it checks that mTexture is not null.

This is to fix this crash that has been happening out in the wild:

0x01444822 	[libGLESv2.dll 	- texture.cpp:1417] 	gl::Texture2D::getRenderTarget(unsigned int)
0x01445615 	[libGLESv2.dll 	- texture.cpp:2048] 	gl::Texture::TextureColorbufferProxy::getRenderTarget()
0x01453bb0 	[libGLESv2.dll 	- context.cpp:1579] 	gl::Context::applyRenderTarget(bool)
0x01455395 	[libGLESv2.dll 	- context.cpp:2413] 	gl::Context::clear(unsigned int)
0x014460c4 	[libGLESv2.dll 	- libglesv2.cpp:611] 	glClear
0x01c53b75 	[chrome.dll 	- gles2_cmd_decoder.cc:3021] 	gpu::gles2::GLES2DecoderImpl::ClearUnclearedRenderbuffers(unsigned int,gpu::gles2::FramebufferManager::FramebufferInfo *)
0x01c53cee 	[chrome.dll 	- gles2_cmd_decoder.cc:3078] 	gpu::gles2::GLES2DecoderImpl::DoFramebufferTexture2D(unsigned int,unsigned int,unsigned int,unsigned int,int)
0x01c57c7d 	[chrome.dll 	- gles2_cmd_decoder_autogen.h:756] 	gpu::gles2::GLES2DecoderImpl::HandleFramebufferTexture2D(unsigned int,gpu::gles2::FramebufferTexture2D const &)
0x01c52197 	[chrome.dll 	- gles2_cmd_decoder.cc:2214] 	gpu::gles2::GLES2DecoderImpl::DoCommand(unsigned int,unsigned int,void const *)
0x01c4fa78 	[chrome.dll 	- cmd_parser.cc:56] 	gpu::CommandParser::ProcessCommand()
0x01c4f8ba 	[chrome.dll 	- gpu_processor.cc:100] 	gpu::GPUProcessor::ProcessCommands()

It was a Windows XP service pack 3 box with Chrome 7.0.524.0. ANGLE was at r418 in this build. I suspect it attempted to glClear after a device lost, which caused needsRenderTarget to fail.

This will hopefully patch it up so it doesn't crash. We'll need to check we aren't assuming mTexture is not null elsewhere.


Review URL: http://codereview.appspot.com/2233045

git-svn-id: https://angleproject.googlecode.com/svn/trunk@427 736b8ea6-26fd-11df-bfd4-992fa37f6226
  29. 16 Sep, 2010 1 commit
  31. 15 Sep, 2010 5 commits
  33. 14 Sep, 2010 2 commits
  35. 11 Sep, 2010 1 commit