flypig.co.uk

Testing out my overnight build I'm pleased to discover that the infinite recursion is now resolved — no more segfaults — but rendering has still not returned. Nevertheless I should check that the mInternalDrawFB and mInternalReadFB are now being set correctly.

Back on Day 204 I placed a breakpoint on CompositorOGL::CreateContext() for this, stepping through to check the values:

Thread 36 "Compositor" hit Breakpoint 1, mozilla::layers::
    CompositorOGL::CreateContext (this=0x7eac003420)
    at gfx/layers/opengl/CompositorOGL.cpp:223
223     already_AddRefed<mozilla::gl::GLContext> CompositorOGL::CreateContext() 
    {
[...]
(gdb) n
mozilla::layers::CompositorOGL::Initialize (this=0x7eac003420, 
    out_failureReason=0x7fa516f730)
    at gfx/layers/opengl/CompositorOGL.cpp:374
374       mGLContext = CreateContext();
(gdb) n
383       if (!mGLContext) {
(gdb) p mGLContext
$5 = {mRawPtr = 0x7eac109140}
(gdb) p mGLContext.mRawPtr
$6 = (mozilla::gl::GLContext *) 0x7eac109140
[...]
(gdb) p mGLContext.mRawPtr->mScreen.mTuple.mFirstA.mUserDrawFB
$12 = 0
(gdb) p mGLContext.mRawPtr->mScreen.mTuple.mFirstA.mUserReadFB
$13 = 0
(gdb) p mGLContext.mRawPtr->mScreen.mTuple.mFirstA.mInternalDrawFB
$14 = 2
(gdb) p mGLContext.mRawPtr->mScreen.mTuple.mFirstA.mInternalReadFB
$15 = 2
(gdb) 

So I'll do the same on the latest version of the code to see whether we have the difference we need. Here's the output:

mozilla::layers::CompositorOGL::Initialize (this=0x7ee0002f10, 
    out_failureReason=0x7f1f923520)
    at gfx/layers/opengl/CompositorOGL.cpp:391
391         mGLContext = CreateContext();
(gdb)
401       if (!mGLContext) {
(gdb) p mGLContext.mRawPtr->mScreen.mTuple.mFirstA.mUserDrawFB
$1 = 0
(gdb) p mGLContext.mRawPtr->mScreen.mTuple.mFirstA.mUserReadFB
$2 = 0
(gdb) p mGLContext.mRawPtr->mScreen.mTuple.mFirstA.mInternalDrawFB
$3 = 2
(gdb) p mGLContext.mRawPtr->mScreen.mTuple.mFirstA.mInternalReadFB
$4 = 2
(gdb) 

That does look a lot more healthy now. So, with that in order, it's back to stepping through the execution, comparing ESR 78 with ESR 91 again. But that's for today. As you can see, it's only a short one today — the checking took a lot longer than I was expecting — but there's plenty still to do and with any luck I'll find time to pick up the pace a bit tomorrow and over the weekend.

If you'd like to read any of my other gecko diary entries, they're all available on my Gecko-dev Diary page.