Index
> Windows > FASMW on Vista 64 - Not RespondingGoto page Previous 1, 2 |
| Author |
|
|
LocoDelAssembly 11 Oct 2009, 01:38
Excellent, I'll make the formal bug report then.
 |
|||
11 Oct 2009, 01:38 |
|
|
kohlrak 11 Oct 2009, 02:08
What exactly causes the freeze though? If the problem's win64 specific, doesn't this mean it's windows' problem (no reason we can't do something about it until they do fix it though)?
|
|||
|
11 Oct 2009, 02:08 |
|
|
LocoDelAssembly 11 Oct 2009, 03:01
I don't know with certainty but the problem may have something to do with this API that doesn't like to have calls to ShowCaret from a WindowProc when WM_OBJECT message is sent. Additionally it seems to be ShowCaret who triggers the WM_OBJECT message and since it is "replied" by calling ShowCaret again then the app gets blocked.
|
|||
|
11 Oct 2009, 03:01 |
|
|
kohlrak 11 Oct 2009, 03:16
Take note of the "type" displayed... Is this then something that would be specific to tablet PCs? I'm still not sure i'm getting this (windows is so complex the developers admit they don't even know what's going on when they're making it).
|
|||
|
11 Oct 2009, 03:16 |
|
|
Tomasz Grysztar 11 Oct 2009, 09:07
In fact, this is not the first time that someone reported problem with fasmw on the machine with tablet attached, so perhaps it's related.
I'm sorry I could not respond to this thread earlier, this was a very busy week for me. I have updated the fasmw package for 1.69.08 core with a fix that gets rid of excessive ShowCaret calling. Please test it. |
|||
|
11 Oct 2009, 09:07 |
|
|
Suzume 11 Oct 2009, 09:45
I'm going to test the new version right away.
Tablet PC ... something about this drifts at the far horizon of my mind. "Vista is somehow making a tablet pc out of a notebook as soon as a Wacom graphics tablet is intalled..." I don't know if I recall correctly here. So that may be relevant - programmers rarely have graphics tablets attached to their notebooks which may be why this behaviour has not been encountered earlier. I do, though I didn't have it connected when I tried to use fasmw. edit: Are you sure that you updated the 16908 zip for windows? The fasmw.exe is flagged as being last modfied on 04. Octobre 2009. The only file from 11th I found at a quick glance was WIN32AX.INC... and compiling the SOURCE into binaries via fasm.exe to a new fasmw.exe didn't show any changes - the IDE still rejects to respond in time. edit2: Here's another dump file created from the fasmw.exe that I assembled from the source code which ships with 16908: http://rapidshare.com/files/291593101/FASMW_dump2.7z.html MD5: A0093591F50543144A0C4154ED57FFEC |
|||
|
11 Oct 2009, 09:45 |
|
|
LocoDelAssembly 11 Oct 2009, 17:09
Suzume, it is like you said, the new FASMW was not included in the update. I gave a look at your dump anyway. The error is the same but the stack has a difference in the last line:
Code: 000bb6a4 74ba4ee2 000608fc 00000000 000bb6d0 oleacc!LookupWindowClass+0x73 Code: 000bada4 74ba4100 00150740 fffffffc 74ba22a8 oleacc!ORIGINAL_AccessibleObjectFromWindow+0x71 But perhaps it is not true because I had the message "*** WARNING: symbols timestamp is wrong 0x4791ad89 0x4791a74d for oleacc.dll" in the third line and that may be doing some mess in the interpretation of the symbols. Just in case I'm missing something I copy it completely: Code: 0:000:x86> kb ChildEBP RetAddr Args to Child 000bac74 76e5d530 000608fc 0000003d ffffffff user32!NtUserMessageCall+0x15 000bacb0 76e5d568 000608fc 0000003d ffffffff user32!SendMessageTimeoutWorker+0xa6 *** WARNING: symbols timestamp is wrong 0x4791ad89 0x4791a74d for oleacc.dll 000bacd8 74ba41c5 000608fc 0000003d ffffffff user32!SendMessageTimeoutW+0x21 000bad04 74ba4100 000608fc fffffffc 74ba22a8 oleacc!ORIGINAL_AccessibleObjectFromWindow+0x71 000bad28 74bc8bb1 000608fc fffffffc 74ba22a8 oleacc!AccessibleObjectFromWindow+0x1a 000bad64 732f2884 000608fc fffffffc 00000000 oleacc!AccessibleObjectFromEvent+0x5c 000badd8 732feaa9 000608fc fffffffc 00000000 tiptsf!CARET::UpdateMSAAEditFieldState+0x37 000bb050 732f1bad 00000000 000608fc 00008002 tiptsf!CARET::_ProcessCaretEvents+0x82b 000bb080 76e783ec 00010100 00008002 000608fc tiptsf!CARET::ProcessCaretEvents+0x8c 000bb0c0 772ee476 000bb0d8 00000000 000bb42c user32!__ClientCallWinEventProc+0x2e *** ERROR: Module load completed but symbols could not be loaded for fasmw.exe 000bb0f4 0040b2b3 000608fc 00000000 0040b000 ntdll_772d0000!KiUserCallbackDispatcher+0x2e WARNING: Stack unwind information not available. Following frames may be wrong. 000bb398 76e58817 000608fc 0000003d 00000000 fasmw+0xb2b3 000bb3c4 76e5898e 0040b000 000608fc 0000003d user32!InternalCallWinProc+0x23 000bb43c 76e59d14 00000000 0040b000 000608fc user32!UserCallWinProcCheckWow+0x109 000bb498 76e59d85 00b9de10 00000000 0000003d user32!DispatchClientMessage+0xe0 000bb4d4 772ee476 000bb4ec 00000000 000bb934 user32!__fnDWORD+0x2b 000bb508 76e5d530 000608fc 0000003d 00000000 ntdll_772d0000!KiUserCallbackDispatcher+0x2e 000bb544 76e5d568 000608fc 0000003d 00000000 user32!SendMessageTimeoutWorker+0xa6 000bb56c 74ba4fa1 000608fc 0000003d 00000000 user32!SendMessageTimeoutW+0x21 000bb6a4 74ba4ee2 000608fc 00000000 000bb6d0 oleacc!LookupWindowClass+0x73 |
|||
|
11 Oct 2009, 17:09 |
|
|
Tomasz Grysztar 11 Oct 2009, 21:07
Sorry, that was my mistake during the packaging. The correct updated .zip is uploaded now.
|
|||
|
11 Oct 2009, 21:07 |
|
|
Suzume 11 Oct 2009, 22:07
It works for me now. Thank you!
 |
|||
|
11 Oct 2009, 22:07 |
|
|
r22 12 Oct 2009, 18:03
Quote: How many ShowCaret's does it take to freeze Vista 64? Just 1! (with WM recursion) |
|||
|
12 Oct 2009, 18:03 |
|
|
Suzume 12 Oct 2009, 22:16
Haha, nice... I'll pin that joke to my desktop background.
 |
|||
|
12 Oct 2009, 22:16 |
|
| Goto page Previous 1, 2 < Last Thread | Next Thread > |
Forum Rules:
|