http://bugs.winehq.com/show_bug.cgi?id=1165
------- Additional Comments From 7ownq0k402(a)sneakemail.com 2002-11-27 06:40 -------
In the bug report, I didn't complete this paragraph:
I am happy to provide debug reports, but I'll need guidance as to how to
proceed, because all debug logs I've created are large, and Photoshop must run
at a reasonable rate so I can kill it before it consumes all memory.
DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
<http://bugs.winehq.com/show_bug.cgi?id=1165>.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND
INSERTED IN THE BUG DATABASE.
http://bugs.winehq.com/show_bug.cgi?id=1165
Summary: Photoshop6 hangs, can potentially lock X server, and
machine.
Product: Wine
Version: unspecified
Platform: PC
OS/Version: Linux
Status: UNCONFIRMED
Severity: critical
Priority: P2
Component: wine-misc
AssignedTo: wine-bugs(a)winehq.com
ReportedBy: 7ownq0k402(a)sneakemail.com
Wine build: Build 20021031 (Not in version list)
Steps to reproduce:
1. Run Photoshop6
2. Photoshop starts running.
3. Photoshop main window comes up (sometimes badly drawn)
4. The splash screen comes up.
5. Towards the end of the load ("Initialising..." is in splash screen) the main
window disappears, as does the splash screen.
6. The whole screen is now locked apart from the KDE panel, with the cursor
showing the hourglass.
7. When menus are lauched from the panel. the mouse works properly in menu.
Menu is draw correctly, but when leave menu, nothing is re-drawn.
8. Sometimes (especially if left in locked state) Photoshop 6 will use all
available memory (real & swap), and linux will hang.
Repeatability:
Always.
Workarounds attempted (None succeded):
1. Attempted using the following DLLs native:
shell,shell32,commtrl,comctl32,commdlg,comdlg32,shlwapi,
ole32,rpcrt4,shlwapi,oleaut32,msvcrt20,msvcrt,msvcp60,shfolder
2. Removed the native wintab32.dll
3. Tried with "Synchronous" = "Y"
This improved the drawing, but didn't fix the bug.
4. Tried with OS set to Win 98.
Wine config:
Is based upon a win98 install.
Photoshop6 was installed in Win98.
Options for exiting lock:
1. Go to a console (CTRL-ALT-FNKEY), and kill wineserver, and any wine apps.
2. Shutdown computer (sometimes possible from KDE panel menu.
3. Reset computer.
Looks almost as if a X grab has occured, on the whole screen apart
from the KDE panel.
I am happy to provide debug reports, but I'll need guidance as to how to
proceed, because all debug logs I've created are large, and Photoshop must run
at a reasonable rate so I can kill it before
Something to do with shfolder.dll?????
Seems to occur near wher dll is loaded.
DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
<http://bugs.winehq.com/show_bug.cgi?id=1165>.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND
INSERTED IN THE BUG DATABASE.
http://bugs.winehq.com/show_bug.cgi?id=1129
------- Additional Comments From andi(a)rhlx01.fht-esslingen.de 2002-11-27 00:48 -------
VFAT_IOCTL_READDIR_BOTH doesn't matter, since it's on a non-VFAT partition
and we just probe for VFAT properties.
Apart from that, no idea.
DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
<http://bugs.winehq.com/show_bug.cgi?id=1129>.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND
INSERTED IN THE BUG DATABASE.
http://bugs.winehq.com/show_bug.cgi?id=1129
------- Additional Comments From tony_lambregts(a)telusplanet.net 2002-11-26 19:12 -------
I would suspect that its something to do with konqueror. Lets see what I can do
<Klaus's comments>
I was just wondering if it may be that wine does not load the bitmaps of the
game. I did an strace and I don't find anything about loading the files
"./dink/Tiles/*.bmp".The only thing I discovered was
open("/usr/share/wine/drivec/Program Files/Dink Smallwood/dink/Tiles",
O_RDONLY|O_LARGEFILE) = 15
ioctl(15, VFAT_IOCTL_READDIR_BOTH, 0x406d16a4) = -1 ENOTTY (Inappropriate
ioctlfor device)
close(15)= 0
open("/usr/share/wine/drivec/Program Files/Dink Smallwood/dink/Tiles",
O_RDONLY|O_NONBLOCK|O_LARGEFILE|O_DIRECTORY) = 15
fstat64(15, {st_mode=S_IFDIR|0777, st_size=1456, ...}) = 0
fcntl64(15, F_SETFD, FD_CLOEXEC) = 0
getdents64(0xf, 0x80c8040, 0x1000, 0x19) = 1904
getdents64(0xf, 0x80c8040, 0x1000, 0x19) = 0
close(15) = 0
Maybe someone smarter than me could tell me what this means.Anyway if I use
"wine -debugmsg +file dink.exe" I find lines saying:
trace:file:CreateFileW L"tiles\\TS01.bmp" GENERIC_READ FILE_SHARE_READ
FILE_SHARE_WRITE OPEN_EXISTING attributes 0x80
trace:file:CreateFileW returning 0x58
trace:file:_lopen ('tiles\TS01.BMP',0000)
trace:file:CreateFileW L"tiles\\TS01.BMP" GENERIC_READ FILE_SHARE_READ
FILE_SHARE_WRITE OPEN_EXISTING attributes 0x0
trace:file:CreateFileW returning 0x58
trace:file:ReadFile 0x58 0x406d2690 14 0x406d264c (nil)
trace:file:ReadFile 0x58 0x406d2668 40 0x406d264c (nil)
trace:file:ReadFile 0x58 0x406d26a4 1024 0x406d264c (nil)
trace:file:_lclose handle 88
</Klaus's comments>
DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
<http://bugs.winehq.com/show_bug.cgi?id=1129>.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND
INSERTED IN THE BUG DATABASE.
http://bugs.winehq.com/show_bug.cgi?id=1129
------- Additional Comments From kniederk(a)math.uni-koeln.de 2002-11-26 18:44 -------
Me again. Sorry for the strange formatting of my emails.I don't know why this happens. I'm using konqueror-2.2.xI was just wondering if it may be that wine does not load the bitmapsof the game. I did an strace and I don't find anything about loadingthe files "./dink/Tiles/*.bmp".The only thing I discovered wasopen("/usr/share/wine/drivec/Program Files/Dink Smallwood/dink/Tiles", O_RDONLY|O_LARGEFILE) = 15ioctl(15, VFAT_IOCTL_READDIR_BOTH, 0x406d16a4) = -1 ENOTTY (Inappropriate ioctlfor device)close(15) = 0open("/usr/share/wine/drivec/Program Files/Dink Smallwood/dink/Tiles", O_RDONLY|O_NONBLOCK|O_LARGEFILE|O_DIRECTORY) = 15fstat64(15, {st_mode=S_IFDIR|0777, st_size=1456, ...}) = 0fcntl64(15, F_SETFD, FD_CLOEXEC) = 0getdents64(0xf, 0x80c8040, 0x1000, 0x19) = 1904getdents64(0xf, 0x80c8040, 0x1000, 0x19) = 0close(15) = 0Maybe someone smarter than me could tell me what this means.Anyway if I use "wine -debugmsg +file dink.exe" I find lines saying:trace:file:CreateFileW L"tiles\\TS01.bmp" GENERIC_READ FILE_SHARE_READ FILE_SHARE_WRITE OPEN_EXISTING attributes 0x80trace:file:CreateFileW returning 0x58trace:file:_lopen ('tiles\TS01.BMP',0000)trace:file:CreateFileW L"tiles\\TS01.BMP" GENERIC_READ FILE_SHARE_READ FILE_SHARE_WRITE OPEN_EXISTING attributes 0x0trace:file:CreateFileW returning 0x58trace:file:ReadFile 0x58 0x406d2690 14 0x406d264c (nil)trace:file:ReadFile 0x58 0x406d2668 40 0x406d264c (nil)trace:file:ReadFile 0x58 0x406d26a4 1024 0x406d264c (nil)trace:file:_lclose handle 88I hope someone can read this and the formatting is not destroyed againthis time.Bye Klaus
DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
<http://bugs.winehq.com/show_bug.cgi?id=1129>.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND
INSERTED IN THE BUG DATABASE.
http://bugs.winehq.com/show_bug.cgi?id=1153
wilson(a)afn.org changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|RESOLVED |CLOSED
------- Additional Comments From wilson(a)afn.org 2002-11-25 20:13 -------
Uwe asked that I close this bug, not merely mark it "Fixed".
DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
<http://bugs.winehq.com/show_bug.cgi?id=1153>.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND
INSERTED IN THE BUG DATABASE.
http://bugs.winehq.com/show_bug.cgi?id=1153
wilson(a)afn.org changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|NEW |RESOLVED
Resolution| |FIXED
------- Additional Comments From wilson(a)afn.org 2002-11-25 20:11 -------
Uwe's patches appear to fix my test application. Thanks all.
DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
<http://bugs.winehq.com/show_bug.cgi?id=1153>.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND
INSERTED IN THE BUG DATABASE.
http://bugs.winehq.com/show_bug.cgi?id=1162
------- Additional Comments From tony_lambregts(a)telusplanet.net 2002-11-25 19:01 -------
I snipped down the log some on my own and ran across this section as I was doing so
0874e8e0:Call USER.106: GETKEYSTATE(0070) ret=029f:0a36 ds=0397
trace:msg:GetKeyState key (0x50) -> 0
0874e8e0:Ret USER.106: GETKEYSTATE() retval=0000 ret=029f:0a36 ds=0397
Now ascii(70) is lower case p and ascii (50) is upper case P. After playing
around a bit I found that GetKeyState is defined in /windows/message.c and this
function converts the keypressed to upper for whatever reason so I commented it
out as in the following patch and now ctrl-P works.
Index: message.c
===================================================================
RCS file: /home/wine/wine/windows/message.c,v
retrieving revision 1.148
diff -u -r1.148 message.c
--- message.c 30 Oct 2002 23:45:38 -0000 1.148
+++ message.c 26 Nov 2002 00:40:19 -0000
@@ -665,7 +665,7 @@
{
INT retval;
- if (vkey >= 'a' && vkey <= 'z') vkey += 'A' - 'a';
+/* if (vkey >= 'a' && vkey <= 'z') vkey += 'A' - 'a'; */
retval = ((WORD)(QueueKeyStateTable[vkey] & 0x80) << 8 ) |
(QueueKeyStateTable[vkey] & 0x80) |
(QueueKeyStateTable[vkey] & 0x01);
Now I suspect that this is not the "Right Fix" tm but I am going to submit it to
wine-devel anyway.
DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
<http://bugs.winehq.com/show_bug.cgi?id=1162>.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND
INSERTED IN THE BUG DATABASE.
http://bugs.winehq.com/show_bug.cgi?id=1122
------- Additional Comments From frick(a)sc-networks.de 2002-11-25 16:41 -------
Well now i can say it definetly - this is the patch:
http://www.winehq.com/hypermail/wine-cvs/2002/10/0469.html
I did an cvs up yesterday and the bug was _not_ gone - assuming the patch
Andreas mentioned was already in this version i used.
DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
<http://bugs.winehq.com/show_bug.cgi?id=1122>.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND
INSERTED IN THE BUG DATABASE.