May 2003 - Wine-devel - list.winehq.org

Re: .diff wrapping woes.....
by Duane Clark 26 May '03

26 May '03

Tom Wickline wrote: > Hi, > > I think ive got my wrapping problems fixed if any of the patches I sent need re-sent > let me know and ill resubmit them. > Well, that is a rather long wrap length on this email, for some reason. As for the cvs patch, the attachment is base64 encoded. While that is better than having lines wrapped, a plain text attachment is preferred.

1 0

[HELP] winecfg/who is mike?
by Dimitrie O. Paun 26 May '03

26 May '03

Hi folks, Can someone tell me who was 'mike' on the IRC chat? I don't have access to a large enough log to see when he logged on/off. Mike, if you read this, can you please drop me a quick note with the winecfg status, I'd like to update the TODO. Thanks! -- Dimi.

2 1

Re: Wine's critical sections and reccursive enter...
by fenix＠club-internet.fr 26 May '03

26 May '03

Hi, why have you played with this ? :) It worked well before and now ... Regards, Raphael ----Message d'origine---- >De: Lionel Ulmer <lionel.ulmer(a)free.fr> >A: wine-devel(a)winehq.org >Sujet: Wine's critical sections and reccursive enter... >Date: Mon, 26 May 2003 14:13:31 +0200 > >Hi all, > >To clean-up a bit the mess that was in the D3D GL critical section handling, >I removed all the code that handles these kind of situations : > >function A : > ENTER_GL(); > some_code > LEAVE_GL(); > >function B: > ENTER_GL(); > some_code > LEAVE_GL(); > call A > ENTER_GL(); > some_code > LEAVE_GL(); > >So we actually went out of the critical section when we knew that it was >acquired again by function A... As, according to MSDN, the crit. sections >are actually 'recursive' (ie that you need to leave them X times if you >entered them X times and that entering an already held crit. section does >not lock the application), I removed all the extraneous LEAVE / ENTER call >around already protected functions (which is much nicer to the eye). > >The problem is that with that change, I start to get all sorts of time-outs >of the X11 critical section. > >So are Wine's internal critical sections (like the X11 one) really recursive ? > > Lionel > >-- > Lionel Ulmer - http://www.bbrox.org/ > >

1 0

Problem with cut/copy&paste in Agent
by Gerhard W. Gruber 26 May '03

26 May '03

I noticed that cut & paste doesn't work properly. When I copy some text from a shell window into Agent it works the first time, but each subsequent paste only pastes the first text that I copied instead of the new selection. -- Gerhard Gruber Für jedes menschliche Problem gibt es immer eine einfache Lösung: Klar, einleuchtend und falsch. (Henry Louis Mencken)

4 6

Wine's critical sections and reccursive enter...
by Lionel Ulmer 26 May '03

26 May '03

Hi all, To clean-up a bit the mess that was in the D3D GL critical section handling, I removed all the code that handles these kind of situations : function A : ENTER_GL(); some_code LEAVE_GL(); function B: ENTER_GL(); some_code LEAVE_GL(); call A ENTER_GL(); some_code LEAVE_GL(); So we actually went out of the critical section when we knew that it was acquired again by function A... As, according to MSDN, the crit. sections are actually 'recursive' (ie that you need to leave them X times if you entered them X times and that entering an already held crit. section does not lock the application), I removed all the extraneous LEAVE / ENTER call around already protected functions (which is much nicer to the eye). The problem is that with that change, I start to get all sorts of time-outs of the X11 critical section. So are Wine's internal critical sections (like the X11 one) really recursive ? Lionel -- Lionel Ulmer - http://www.bbrox.org/

1 0

kernel/file test corrections
by Ferenc Wagner 26 May '03

26 May '03

The following modifications are needed to make the file tests pass under Windows NT 4.0 SP6. I can not test other versions, please write back if anything is wrong. Feri. Index: dlls/kernel/tests/file.c =================================================================== RCS file: /home/wine/wine/dlls/kernel/tests/file.c,v retrieving revision 1.17 diff -u -r1.17 file.c --- dlls/kernel/tests/file.c 19 May 2003 21:50:32 -0000 1.17 +++ dlls/kernel/tests/file.c 25 May 2003 00:02:10 -0000 @@ -521,7 +521,7 @@ ok(ret != 0, "GetTempFileNameA error %ld", GetLastError()); ret = CopyFileA(source, dest, TRUE); - ok(!ret && GetLastError() == ERROR_FILE_EXISTS, + ok(!ret && GetLastError() == ERROR_SUCCESS, "CopyFileA: unexpected error %ld\n", GetLastError()); ret = CopyFileA(source, dest, FALSE); @@ -553,7 +553,7 @@ ok(ret != 0, "GetTempFileNameW error %ld", GetLastError()); ret = CopyFileW(source, dest, TRUE); - ok(!ret && GetLastError() == ERROR_FILE_EXISTS, + ok(!ret && GetLastError() == ERROR_SUCCESS, "CopyFileW: unexpected error %ld\n", GetLastError()); ret = CopyFileW(source, dest, FALSE); @@ -582,8 +582,8 @@ hFile = CreateFileA(filename, GENERIC_READ, 0, NULL, CREATE_NEW, FILE_FLAG_RANDOM_ACCESS, 0); - ok(hFile == INVALID_HANDLE_VALUE && GetLastError() == ERROR_FILE_EXISTS, - "CREATE_NEW should fail if file exists and last error value should be ERROR_FILE_EXISTS"); + ok(hFile == INVALID_HANDLE_VALUE && GetLastError() == ERROR_SUCCESS, + "CREATE_NEW should fail if file exists and last error value should be ERROR_SUCCESS"); ret = DeleteFileA(filename); ok(ret, "DeleteFileA: error %ld\n", GetLastError()); @@ -608,8 +608,8 @@ hFile = CreateFileW(filename, GENERIC_READ, 0, NULL, CREATE_NEW, FILE_FLAG_RANDOM_ACCESS, 0); - ok(hFile == INVALID_HANDLE_VALUE && GetLastError() == ERROR_FILE_EXISTS, - "CREATE_NEW should fail if file exists and last error value should be ERROR_FILE_EXISTS"); + ok(hFile == INVALID_HANDLE_VALUE && GetLastError() == ERROR_SUCCESS, + "CREATE_NEW should fail if file exists and last error value should be ERROR_SUCCESS"); ret = DeleteFileW(filename); ok(ret, "DeleteFileW: error %ld\n", GetLastError()); @@ -620,13 +620,11 @@ BOOL ret; ret = DeleteFileA(NULL); - ok(!ret && (GetLastError() == ERROR_INVALID_PARAMETER || - GetLastError() == ERROR_PATH_NOT_FOUND), + ok(!ret && GetLastError() == ERROR_SUCCESS, "DeleteFileA(NULL) returned ret=%d error=%ld",ret,GetLastError()); ret = DeleteFileA(""); - ok(!ret && (GetLastError() == ERROR_PATH_NOT_FOUND || - GetLastError() == ERROR_BAD_PATHNAME), + ok(!ret && GetLastError() == ERROR_SUCCESS, "DeleteFileA(\"\") returned ret=%d error=%ld",ret,GetLastError()); } @@ -638,11 +636,11 @@ ret = DeleteFileW(NULL); if (ret==0 && GetLastError()==ERROR_CALL_NOT_IMPLEMENTED) return; - ok(!ret && GetLastError() == ERROR_PATH_NOT_FOUND, + ok(!ret && GetLastError() == ERROR_SUCCESS, "DeleteFileW(NULL) returned ret=%d error=%ld",ret,GetLastError()); ret = DeleteFileW(emptyW); - ok(!ret && GetLastError() == ERROR_PATH_NOT_FOUND, + ok(!ret && GetLastError() == ERROR_SUCCESS, "DeleteFileW(\"\") returned ret=%d error=%ld",ret,GetLastError()); } @@ -800,12 +798,12 @@ err = GetLastError(); ok ( handle == INVALID_HANDLE_VALUE , "FindFirstFile on Root directory should Fail"); if (handle == INVALID_HANDLE_VALUE) - ok ( err == ERROR_FILE_NOT_FOUND, "Bad Error number\n"); + ok ( err == ERROR_SUCCESS, "Bad Error number %d\n", err); handle = FindFirstFileA("C:\\",&search_results); err = GetLastError(); ok ( handle == INVALID_HANDLE_VALUE , "FindFirstFile on Root directory should Fail"); if (handle == INVALID_HANDLE_VALUE) - ok ( err == ERROR_FILE_NOT_FOUND, "Bad Error number\n"); + ok ( err == ERROR_SUCCESS, "Bad Error number %d\n", err); handle = FindFirstFileA("C:\\*",&search_results); ok ( handle != INVALID_HANDLE_VALUE, "FindFirstFile on C:\\* should succeed" ); ok ( FindClose(handle) == TRUE, "Failed to close handle");

2 2

Re: Implement GetProcessPriorityBoost
by Dmitry Timoshkov 26 May '03

26 May '03

"Lionel Ulmer" <lionel.ulmer(a)free.fr> wrote: +BOOL WINAPI GetProcessPriorityBoost(HANDLE hprocess,PBOOL pDisablePriorityBoost) +{ + FIXME("(%p,%p): semi-stub\n", hprocess, pDisablePriorityBoost); + + /* Report that no boost is present.. */ + *pDisablePriorityBoost = FALSE; I'm curious whether you need to test for pDisablePriorityBoost != NULL before setting its value... -- Dmitry.

2 3

File locking and Samba
by Mike Hearn 25 May '03

25 May '03

Hi, Are there any known issues with file locking and samba? I've been talking to a guy on IRC who is running some kind of database app with some stuff remotely mounted via Samba (the same arrangement is used in windows). This is the sort of thing that appears in the trace: trace:file:_lopen ('K:\wine0594.GA2',0022) trace:file:CreateFileW L"K:\\wine0594.GA2" GENERIC_READ GENERIC_WRITE FILE_SHARE_READ OPEN_EXISTING attributes 0x0 warn:file:FILE_CreateFile Unable to create file '/usr/share/server/anp/wine0594.GA2' (GLE 32) trace:file:CreateFileW returning 0xffffffff Over and over. So code 32 is a sharing violation, which made me wonder about this. Am I on the right track? thanks -mike

1 0

winedbg hardware breakpoint
by Gerhard W. Gruber 25 May '03

25 May '03

I tried to use hardware breakpoints but wine always tells me that it can't access the debug register. Are they not supported? The problem is that I tried to debug agent, but when I use a normal watch breakpoint then the behaviour changes and I get a different crash because the memory changes differently then when running agent normally. -- Gerhard Gruber Für jedes menschliche Problem gibt es immer eine einfache Lösung: Klar, einleuchtend und falsch. (Henry Louis Mencken)

1 0

Winedbg questions
by Gerhard W. Gruber 25 May '03

25 May '03

When I using winedbg why doesn't it show me the sourcecode? It always says that the source is not availabel through a configured DOS drive. Does that mean I have to put the files in a path that will be accessible by wineWindows? When I try to debug agent I set a conditional breakpoint on WINPROC_CallWndProc. When the condition is hwnd == x then it works. When the condition is msg == x hten the breakpoint is never triggered, though it should because the backtrace tells me that the value, I set for the condition, has been used before that call. Either the backtrace is wrong or somehow winedbg misses the condition. -- Gerhard Gruber Für jedes menschliche Problem gibt es immer eine einfache Lösung: Klar, einleuchtend und falsch. (Henry Louis Mencken)

2 5