More deadlocks + backtrace

[Shachar Shemesh]

I ran "spyxx" from MSVC on my deadlocking program, and everything was 
ok. I then quite the program itself (Microsoft Digital Image Pro was run 
with +relay, spy was run without any debug values). I left spy running 
for a while (~2 days), as I had other things to do, and I wanted the 
window handles it collected.

At some point, the console that launched spy began showing the following:
$ wine SPYXX
err:font:ReadFontDir Can't open directory "/win/windows/fonts"
fixme:mdi:MDIRefreshMenu partially function stub
fixme:mdi:MDIRefreshMenu partially function stub
err:toolbar:TOOLBAR_NotifyFormat wrong response to WM_NOTIFYFORMAT (0), 
assuming ANSI
err:win:CreateWindowExA bad class name "msctls_updown"
err:ntdll:RtlpWaitForCriticalSection section 0x40380d68 "?" wait timed 
out in thread 0011,
 blocked by 0000, retrying (60 sec)
err:ntdll:RtlpWaitForCriticalSection section 0x40380d68 "?" wait timed 
out in thread 0011,
 blocked by 0000, retrying (60 sec)
err:ntdll:RtlpWaitForCriticalSection section 0x40380d68 "?" wait timed 
out in thread 0011,
 blocked by 0000, retrying (60 sec)
err:ntdll:RtlpWaitForCriticalSection section 0x40380d68 "?" wait timed 
out in thread 0011,
 blocked by 0000, retrying (60 sec)
err:ntdll:RtlpWaitForCriticalSection section 0x40380d68 "?" wait timed 
out in thread 0011,
 blocked by 0000, retrying (60 sec)

Here are the backtraces:
Wine-dbg>walk process
 pid      threads  parent   executable
> 00000010 2        00000000 'Y:\sources\Transfer\RemoteDebug\SPYXX.EXE'

Wine-dbg>walk thread
process  tid      prio
00000010 (D) Y:\sources\Transfer\RemoteDebug\SPYXX.EXE
        00000012    0 <==
        00000011    0
Wine-dbg>bt 0x11
Backtrace:
=>0 0x4010e8ab (NTDLL.DLL._memccpy+0x5202b in libc.so.6) (ebp=40032c60)
  1 0x401f4002 (NTDLL_wait_for_multiple_objects+0xbf(count=0x0, 
handles=0x0, flags=0x8, timeout=0x40032d10) [sync.c:578] in NTDLL.DLL) 
(ebp=40032cf8)
  2 0x401f1edd (usr1_handler+0x4e(__signal=0xa, __context=0x8f) 
[signal_i386.c:1123] in NTDLL.DLL) (ebp=40032d1c)
  3 0x4006e498 (NTDLL.DLL.toupper+0x6208 in libc.so.6) (ebp=4080f7e0)
  4 0x401f4002 (NTDLL_wait_for_multiple_objects+0xbf(count=0x1, 
handles=0x4080f8cc, flags=0xc, timeout=0x4080f900) [sync.c:578] in 
NTDLL.DLL) (ebp=4080f878)
  5 0x401f40ae (NTDLL.DLL.NtWaitForMultipleObjects+0x74 in NTDLL.DLL) 
(ebp=4080f8a0)
  6 0x401f40fe (NtWaitForSingleObject+0x42(handle=0x88, alertable=0x0, 
timeout=0x4080f900) [sync.c:611] in NTDLL.DLL) (ebp=4080f8c4)
  7 0x401cda35 (RtlpWaitForCriticalSection+0x127(crit=0x40380d68) 
[critsection.c:193] in NTDLL.DLL) (ebp=4080f964)
  8 0x401cdc25 (RtlEnterCriticalSection+0x51(crit=0x40380d68) 
[critsection.c:255] in NTDLL.DLL) (ebp=4080f97c)
  9 0x404f2555 (wine_pthread_mutex_lock+0x34(mutex=0x101593c8) 
[pthread.c:182] in KERNEL32.DLL) (ebp=4080f98c)
  10 0x0011273a (__pthread_mutex_lock+0x25(mutex=0x101593c8) 
[kthread.c:588] in wine-kthread) (ebp=4080f99c)
  11 0x40f409ea (_end+0x75bb6 in libX11.so.6) (ebp=4080f9bc)
  12 0x40f0d799 (_end+0x42965 in libX11.so.6) (ebp=4080fa1c)
  13 0x40e8b39e (X11DRV_InitKeyboard+0x84(key_state_table=0x40b66880) 
[keyboard.c:1332] in X11DRV.DLL) (ebp=4080fadc)
  14 0x40e8c1c4 (X11DRV_MappingNotify+0x35(event=0x4080fb94) 
[keyboard.c:1635] in X11DRV.DLL) (ebp=4080fafc)
  15 0x40e83647 (.L164+0xb [event.c] in X11DRV.DLL) (ebp=4080fb7c)
  16 0x40e82e65 (process_events+0x71(data=0x40380d40) [event.c:143] in 
X11DRV.DLL) (ebp=4080fbf8)
  17 0x40e82ffa (X11DRV_MsgWaitForMultipleObjectsEx+0x17b(count=0x1, 
handles=0x40380c6c, timeout=0xffffffff, mask=0x0, flags=0x0) 
[event.c:184] in X11DRV.DLL) (ebp=4080fd38)
  18 0x40ad5f96 (GetMessageW+0x287(msg=0x46becc, hwnd=0x0, first=0x0, 
last=0x0) [message.c:2363] in USER32.DLL) (ebp=4080fdf4)
  19 0x40ad6020 (GetMessageA+0x33(msg=0x46becc, hwnd=0x0, first=0x0, 
last=0x0) [message.c:2380] in USER32.DLL) (ebp=4080fe14)
  20 0x6c3be6e9 (MFC42.DLL.5307+0x13 in MFC42.DLL) (ebp=0046becc)
  21 0x0000000f (ebp=000100d6)

Wine-dbg>bt 0x12
Backtrace:
=>0 0x4010e8ab (NTDLL.DLL._memccpy+0x5202b in libc.so.6) (ebp=4030fc60)
  1 0x401f4002 (NTDLL_wait_for_multiple_objects+0xbf(count=0x0, 
handles=0x0, flags=0x8, timeout=0x4030fd10) [sync.c:578] in NTDLL.DLL) 
(ebp=4030fcf8)
  2 0x401f1edd (usr1_handler+0x4e(__signal=0xa, __context=0x8f) 
[signal_i386.c:1123] in NTDLL.DLL) (ebp=4030fd1c)
  3 0x4006e498 (NTDLL.DLL.toupper+0x6208 in libc.so.6) (ebp=419eb450)
  4 0x401f4002 (NTDLL_wait_for_multiple_objects+0xbf(count=0x1, 
handles=0x419eb53c, flags=0xc, timeout=0x419eb530) [sync.c:578] in 
NTDLL.DLL) (ebp=419eb4e8)
  5 0x401f40ae (NTDLL.DLL.NtWaitForMultipleObjects+0x74 in NTDLL.DLL) 
(ebp=419eb510)
  6 0x404fcb20 (WaitForMultipleObjectsEx+0x1d5(count=0x1, 
handles=0x419eb670, wait_all=0x0, timeout=0x32, alertable=0x0) 
[sync.c:167] in KERNEL32.DLL) (ebp=419eb648)
  7 0x404fc8b8 (WaitForSingleObject+0x3c(handle=0x64, timeout=0x32) 
[sync.c:95] in KERNEL32.DLL) (ebp=419eb668)
  8 0x004232fa (SPYXX.EXE..text+0x222fa in SPYXX.EXE) (ebp=40aab4df)
  9 0x0168ec81 (SPYXX.EXE..rsrc+0x1221c81 in wine-kthread) (ebp=53e58955)
*** Invalid address 0x53e58955 (_end+0x121f453d)

I hope this helps someone.

[Mike Hearn]

There was recently (within the last few days) a commit to CVS by
Alexandre that works around a threading bug in Xlib - how recent was the
build you were using?

[Shachar Shemesh]

Mike Hearn wrote:
> There was recently (within the last few days) a commit to CVS by
> Alexandre that works around a threading bug in Xlib - how recent was the
> build you were using?

Next time - read my mail ;-)

I said spyxx was running for about 48 hours. That is, before Alexandre 
commited that patch. I'm recompiling at the moment to see whether the 
problems stop recurring.

I submitted the backtrace, partly in the hope that someone will say "I 
know this - it's now fixed", and partly so that if it's not fixed (but I 
don't manage to backtrace it), I'll have it on record (following Linus' 
backup methodology).

Shachar

[Mike Hearn]

On Sat, 2003-12-13 at 15:26, Shachar Shemesh wrote:
> Next time - read my mail ;-)
>
> I said spyxx was running for about 48 hours. That is, before Alexandre 
> commited that patch. I'm recompiling at the moment to see whether the 
> problems stop recurring.

Hmm, sorry, I couldn't remember offhand how long ago I saw that commit.
I thought it was more like 4 or 5 days ago, but I could well be losing
track of time at the moment, it wouldn't surprise me :)