Add diagnostic information for CreateDIBSection failures

Add diagnostic information for CreateDIBSection failures

This tries to grab useful diagnostic information from the system when
CreateDIBSection fails since we're about to crash anyway. This tries to categorize
the failure as being out of GDI handles, out of memory generally, or unknown and
puts more information on the stack.

Based on code from cpu@ in r116646 and from vangelis@ in https://codereview.chromium.org/57053002/

R=cpu
TBR=reed
NOTRY=true since nacl_integration is being stupid
BUG=275046
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=235667

[Vangelis Kokkevis, 2013-11-14 01:48:04]

Do you want to incorporate some of the checks for failed object deletions from
https://codereview.chromium.org/57053002 as well in case they point us to
resource leaks?

[jamesr, 2013-11-14 01:53:07]

@cpu - I'm seeing distinct crash instructions in the disassembly.  For instance
here's the two checks:

    int num_gdi_handles = GetGuiResources(GetCurrentProcess(),
                                          GR_GDIOBJECTS);
    if (num_gdi_handles == 0) {
      int get_gui_resources_error = GetLastError();
      base::debug::Alias(&get_gui_resources_error);
      CHECK(false);
    }

    base::debug::Alias(&num_gdi_handles);
    const int kLotsOfHandles = 9990;
    if (num_gdi_handles > kLotsOfHandles)
      CHECK(false);

turn into:

   74 6d6faa1b ff1518c1786d    call    dword ptr [skia!_imp__GetCurrentProcess
(6d78c118)]
   74 6d6faa21 50              push    eax
   74 6d6faa22 ff1540c3786d    call    dword ptr [skia!_imp__GetGuiResources
(6d78c340)]
   75 6d6faa28 8b3dccc3786d    mov     edi,dword ptr
[skia!_imp_??1LogMessageloggingQAEXZ (6d78c3cc)]
   75 6d6faa2e 3bc6            cmp     eax,esi
   75 6d6faa30 8b35c8c3786d    mov     esi,dword ptr
[skia!_imp_??0LogMessageloggingQAEPBDHHZ (6d78c3c8)]
   75 6d6faa36 894590          mov     dword ptr [ebp-70h],eax
   75 6d6faa39 7551            jne     skia!`anonymous
namespace'::CreateHBitmap+0x15c (6d6faa8c)

skia!`anonymous namespace'::CreateHBitmap+0x10b
[d:\chrome\src\skia\ext\bitmap_platform_device_win.cc @ 76]:
   76 6d6faa3b ff1524c1786d    call    dword ptr [skia!_imp__GetLastError
(6d78c124)]
   77 6d6faa41 8d4d94          lea     ecx,[ebp-6Ch]
   77 6d6faa44 51              push    ecx
   77 6d6faa45 894594          mov     dword ptr [ebp-6Ch],eax
   77 6d6faa48 ffd3            call    ebx
   77 6d6faa4a 83c404          add     esp,4
   78 6d6faa4d 6a04            push    4
   78 6d6faa4f 6a4e            push    4Eh
   78 6d6faa51 689c077a6d      push    offset skia!`string' (6d7a079c)
   78 6d6faa56 8d8dd0feffff    lea     ecx,[ebp-130h]
   78 6d6faa5c ffd6            call    esi
   78 6d6faa5e 83c008          add     eax,8
   78 6d6faa61 6884077a6d      push    offset skia!`string' (6d7a0784)
   78 6d6faa66 50              push    eax
   78 6d6faa67 c745fc00000000  mov     dword ptr [ebp-4],0
   78 6d6faa6e c7459801000000  mov     dword ptr [ebp-68h],1
   78 6d6faa75 e846c8ffff      call   
skia!std::operator<<<std::char_traits<char> > (6d6f72c0)
   78 6d6faa7a 83c408          add     esp,8
   78 6d6faa7d 8d8dd0feffff    lea     ecx,[ebp-130h]
   78 6d6faa83 c745fcffffffff  mov     dword ptr [ebp-4],0FFFFFFFFh
   78 6d6faa8a ffd7            call    edi

skia!`anonymous namespace'::CreateHBitmap+0x15c
[d:\chrome\src\skia\ext\bitmap_platform_device_win.cc @ 81]:
   81 6d6faa8c 8d5590          lea     edx,[ebp-70h]
   81 6d6faa8f 52              push    edx
   81 6d6faa90 ffd3            call    ebx
   81 6d6faa92 83c404          add     esp,4
   83 6d6faa95 817d9006270000  cmp     dword ptr [ebp-70h],2706h
   83 6d6faa9c 7e40            jle     skia!`anonymous
namespace'::CreateHBitmap+0x1ae (6d6faade)

skia!`anonymous namespace'::CreateHBitmap+0x16e
[d:\chrome\src\skia\ext\bitmap_platform_device_win.cc @ 84]:
   84 6d6faa9e 6a04            push    4
   84 6d6faaa0 6a54            push    54h
   84 6d6faaa2 689c077a6d      push    offset skia!`string' (6d7a079c)
   84 6d6faaa7 8d8dd0feffff    lea     ecx,[ebp-130h]
   84 6d6faaad ffd6            call    esi
   84 6d6faaaf 83c008          add     eax,8
   84 6d6faab2 6884077a6d      push    offset skia!`string' (6d7a0784)
   84 6d6faab7 50              push    eax
   84 6d6faab8 c745fc01000000  mov     dword ptr [ebp-4],1
   84 6d6faabf c7459802000000  mov     dword ptr [ebp-68h],2
   84 6d6faac6 e8f5c7ffff      call   
skia!std::operator<<<std::char_traits<char> > (6d6f72c0)
   84 6d6faacb 83cbff          or      ebx,0FFFFFFFFh
   84 6d6faace 83c408          add     esp,8
   84 6d6faad1 8d8dd0feffff    lea     ecx,[ebp-130h]
   84 6d6faad7 895dfc          mov     dword ptr [ebp-4],ebx
   84 6d6faada ffd7            call    edi
   84 6d6faadc eb03            jmp     skia!`anonymous
namespace'::CreateHBitmap+0x1b1 (6d6faae1)

so i think 6d6faa8a and 6d6faada are separate crash calls.  Look right?

[jamesr, 2013-11-14 01:54:47]

On 2013/11/14 01:48:04, Vangelis Kokkevis wrote:
> Do you want to incorporate some of the checks for failed object deletions from
> https://codereview.chromium.org/57053002 as well in case they point us to
> resource leaks?

I like those checks to but am leaning towards having those be a separate patch
just so if we want to merge or revert one of the diagnostics it doesn't
interfere with the other.  Right now a CreateDIBSection failure is always a
crash so this patch doesn't add new crashes.  Invalid object/dc management is
probably bad but isn't necessarily a crash today.

[cpu_(ooo_6.6-7.5), 2013-11-14 20:25:53]

lgtm

[commit-bot: I haz the power, 2013-11-14 20:31:58]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/jamesr@chromium.org/70193011/1

[commit-bot: I haz the power, 2013-11-14 22:24:55]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/jamesr@chromium.org/70193011/1

[commit-bot: I haz the power, 2013-11-15 02:15:32]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/jamesr@chromium.org/70193011/1

[commit-bot: I haz the power, 2013-11-15 04:14:29]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/jamesr@chromium.org/70193011/1

[commit-bot: I haz the power, 2013-11-15 07:08:38]

Retried try job too often on win_rel for step(s) nacl_integration
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=win_rel&nu...

[commit-bot: I haz the power, 2013-11-15 08:46:57]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/jamesr@chromium.org/70193011/1

[commit-bot: I haz the power, 2013-11-15 10:32:27]

Retried try job too often on win_rel for step(s) nacl_integration
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=win_rel&nu...

[jamesr, 2013-11-15 21:07:50]

On 2013/11/15 10:32:27, I haz the power (commit-bot) wrote:
> Retried try job too often on win_rel for step(s) nacl_integration
>
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=win_rel&nu...

seriously?

[commit-bot: I haz the power, 2013-11-15 22:20:11]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/jamesr@chromium.org/70193011/1

[commit-bot: I haz the power, 2013-11-16 00:55:37]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/jamesr@chromium.org/70193011/1

[commit-bot: I haz the power, 2013-11-18 08:33:08]

Change committed as 235667