March 2006 - Wine-Devel - list.winehq.org

Race condition in keyboard code?
by Peter Åstrand March 24, 2006

March 24, 2006

Hi, I've been debugging a very strange keyboard problem this afternoon: In some cases the key combination AltGr+2 produces "2" (wrong) rather than "@" (correct), when using a Swedish layout. The problem seems to be a race condition: If a KeyEvent occurs almost directly after the "2" key is pressed, the AltGr state is incorrect. The problem has been verified with this simple test program: /* gcc -Wall xfake.c -o xfake -L/usr/X11R6/lib -lX11 -lXtst */ #include <X11/Xlib.h> #include <X11/extensions/XTest.h> int main( int argc, char* argv[] ) { Display* dpy; dpy = XOpenDisplay( NULL ); XTestFakeKeyEvent(dpy, 0x71, True, CurrentTime); XTestFakeKeyEvent(dpy, 0xb, True, CurrentTime); XSync(dpy, False); //sleep(1); XTestFakeKeyEvent(dpy, 0x71, False, CurrentTime); XTestFakeKeyEvent(dpy, 0xb, False, CurrentTime); XCloseDisplay( dpy ); return 0; } (0x71 is Mode_Switch, 0xb is the "2" key). I'm executing the program while having Wines Notepad active. If the sleep() is commented out, a "2" is inserted into Notepad. If the sleep() is active, "@" is inserted. More evidence: When debugging with WINEDEBUG=keyboard,key, the traces for X11DRV_ToUnicodeEx is printed after the Mode_Switch KeyRelease trace. Now I wonder, is this a problem with: 1) Wine or 2) The Windows API or 3) The way I'm using X11/the keyboard (generating events that quickly). ? I do understand why this happens: X11DRV_ToUnicodeEx uses the global, static variable AltGrMask. Is this a bug? Regards, -- Peter Åstrand ThinLinc Chief Developer Cendio http://www.cendio.se Teknikringen 3 583 30 Linköping Phone: +46-13-21 46 00

2 3

Re: SERVER: Only reset the notify event flag if there's no events pending.
by Alexandre Julliard March 24, 2006

March 24, 2006

Mike McCormack <mike(a)codeweavers.com> writes: > diff --git a/server/change.c b/server/change.c > index 4349d7e..67a47af 100644 > --- a/server/change.c > +++ b/server/change.c > @@ -1130,7 +1130,7 @@ DECL_HANDLER(read_directory_changes) > dir->signaled--; > > /* clear the event */ > - if (event) > + if (event && !dir->signaled) > reset_event( event ); This breaks the tests: change.c:518: Test failed: action wrong change.c:522: Test failed: ov.Internal wrong change.c:523: Test failed: ov.InternalHigh wrong change.c:582: Test failed: should timeout change.c:590: Test failed: ov.Internal wrong change.c:603: Test failed: should timeout change.c:669: Test failed: should timeout make[1]: *** [change.ok] Error 7 -- Alexandre Julliard julliard(a)winehq.org

1 0

Re:wine.inf: Add a shdocvw.dll stub dll to the system directory
by Louis. Lenders March 24, 2006

March 24, 2006

What about adding a few more, like opengl32 (http://bugs.winehq.org/show_bug.cgi?id=3877) and some more (there are more of these bugs in bugzilla that can be solved that way, i could make a quick list of them if necessary) Regards Louis Module: wine Branch: refs/heads/master Commit: c3c112de231fae2f28d3f685c341a204ddef26a9 URL: http://source.winehq.org/git/?p=wine.git;a=commit;h=c3c112de231fae2f28d3f68… Author: Mike McCormack <mike <at> codeweavers.com> Date: Wed Mar 22 23:46:31 2006 +0900 wine.inf: Add a shdocvw.dll stub dll to the system directory. --- tools/wine.inf | 1 + 1 files changed, 1 insertions(+), 0 deletions(-) diff --git a/tools/wine.inf b/tools/wine.inf index 747eb83..17d9144 100644 --- a/tools/wine.inf +++ b/tools/wine.inf <at> <at> -2114,6 +2114,7 <at> <at> HKLM,%CurrentVersion%\Telephony\Country 11,,help.exe,winhelp.exe 11,,msiexec.exe 11,,notepad.exe +11,,shdocvw.dll 11,,progman.exe 11,,regsvr32.exe 11,,wcmd.exe ___________________________________________________________ To help you stay safe and secure online, we've developed the all new Yahoo! Security Centre. http://uk.security.yahoo.com

4 3

Re: [opengl] check drawable and context Visual IDs in wglMakeCurrent()
by Tomas Carnecky March 24, 2006

March 24, 2006

Tomas Carnecky wrote: > Changelog: > - fail if the drawable and context Visual IDs don't match. > > why? because this will produce a BadMatch and crash the application so > instead of the crash rather return GL_FALSE and let the application > handle it. Please don't apply this patch yet. It fixes one case but there is more to be aware of. This BadMatch bug is triggered if an application creates a window (with a default Visual ID) and a custom (non-default) pixel format, the IDs were 0x21 and 0x23 in my case (glxinfo/xdpyinfo report Visual ID in the range from 0x21 to 0x70). But! I've seen that World of Warcraft calls wglMakeCurrent() with a darawable that has Visual ID 0x71 and a context with Visual ID 0x21. Now 0x71 is not defined (according to the glxino output) but it works fine, eg. glXMakeCurrent() doesn't produce the X Error in that case. So I went up the backtrace, into create_glxpixmap(). I let this function fail if it would produce a X Error, eg. if "depth of pixmap does not match the GLX_BUFFER_SIZE value of vis". Then I had to modify wglMakeCurrent() to respect the create_glxpixmap() failure and return FALSE. Works good so far, framerate in WoW went up from ~20 to up to 70 fps, average is somewhere between 40-50 !!! This is incredible. So.. in this attachment you'll find a patch that does what I've just described. I can't test it on anything else than WoW, so if someone would please review it and test with outher opengl/d3d applications it would be great. Also, it would be great if we could put the *Swap*Buffers() into their own log domain, something like 'swapbuffers', because the trace is usually useless, only when you explicitly look whether these functions are called or not, otherwise they only fill the log with garbage. thanks tom

4 6

Known issues area on main wiki page
by Tom Spear (Dustin Booker, Dustin Navea) March 24, 2006

March 24, 2006

I'm not trying to be a whiner here, but I think enough is enough when it comes to having to close bugs as a duplicate. There has to be some simple solution to cut down on the number of new bugs filed that report that winecfg crashes when clicking on the audio tab. I know that the subject above isn't it, but it is at least a start. I was about to create it myself, but my better judgement said I should ask here first. Can someone please(!) put something on some page that almost everyone sees saying that if you are having a problem with winecfg crashing when clicking on the audio tab to check out bug 4051 first? Please! Thanks Tom

13 24

[PATCH] server/mailslot.c doesn't compile on certain systems. [Added ChangeLog]
by Segin March 24, 2006

March 24, 2006

I have noticed that including stdint.h to server/mailslot.c fixes it for me. Those on wine-patches and not wine-devel, please CC any replies. ChangeLog: Includes stdint.h in server/mailslot.c, added comment explaining why. --- server/mailslot.c.orig 2006-03-23 22:37:32.000000000 -0500 +++ server/mailslot.c 2006-03-23 22:40:32.000000000 -0500 @@ -34,6 +34,8 @@ #include <stdlib.h> #include <sys/time.h> #include <sys/types.h> +/* include stdint.h cause of some systems breaking */ +#include <stdint.h> #ifdef HAVE_SYS_IOCTL_H #include <sys/ioctl.h>

1 0

[PATCH] server/mailslot.c doesn't compile on certain systems.
by Segin March 24, 2006

March 24, 2006

I have noticed that including stdint.h to server/mailslot.c fixes it for me. Those on wine-patches and not wine-devel, please CC any replies. --- server/mailslot.c.orig 2006-03-23 22:37:32.000000000 -0500 +++ server/mailslot.c 2006-03-23 22:40:32.000000000 -0500 @@ -34,6 +34,8 @@ #include <stdlib.h> #include <sys/time.h> #include <sys/types.h> +/* include stdint.h cause of some systems breaking */ +#include <stdint.h> #ifdef HAVE_SYS_IOCTL_H #include <sys/ioctl.h>

1 0

rebar patch
by qingdoa daoo March 24, 2006

March 24, 2006

Hi Can i suggest a patch to wine's rebar src The patch is made against current cvs source. [root(a)localhost comctl32]# diff -u rebar.c.cvs rebar.c > attach.diff --- rebar.c.cvs 2006-03-24 09:53:28.000000000 +0800 +++ rebar.c 2006-03-24 09:38:29.000000000 +0800 @@ -2820,7 +2820,7 @@ } if (lprbbi->fMask & RBBIM_SIZE) - lprbbi->cx = lpBand->cx; + lprbbi->cx = lpBand->ccx; if (lprbbi->fMask & RBBIM_BACKGROUND) lprbbi->hbmBack = lpBand->hbmBack; @@ -2902,7 +2902,7 @@ } if (lprbbi->fMask & RBBIM_SIZE) - lprbbi->cx = lpBand->cx; + lprbbi->cx = lpBand->ccx; if (lprbbi->fMask & RBBIM_BACKGROUND) lprbbi->hbmBack = lpBand->hbmBack; It's a fix to the problem that apps can't remember rebar layout across runs. When apps exit, they often ask the rebar for current band info so that the next time they run, they can tell the rebar how to layout the bands. When they ask about the size of a band, current code passes them lpBand->cx, which is a value passed by apps at band creation time, validated and stored by rebar, and then never gets updated. So apps alway get the old size value, while lpBand->ccx as the name suggests is the current rebar size. Tested with some apps, and no regression observed. Source Insight works better, so is Internet Explorer. I have actually made more changes to rebar.c. Wondering if some could review this patch and submit it. So that I can suggest subsequent patches against the updated source. --------------------------------- Blab-away for as little as 1¢/min. Make PC-to-Phone Calls using Yahoo! Messenger with Voice.

2 1

GoogleEarth
by James Courtier-Dutton March 24, 2006

March 24, 2006

Hi, Can one run GoogleEarth under wine? It does not seem to run here under the default settings. James

7 6

build failure on Gentoo Linux, recommended fix included.
by Segin March 24, 2006

March 24, 2006

Kernel 2.6.15-gentoo-r7 (2.6.15.7) glibc 2.3.6 (NTPL/NTPL only) Build error in server/mailslot.c: mailslot.c: In function 'mailslot_next_msg_size': mailslot.c:221: error: 'uint32_t' undeclared (first use in this function) mailslot.c:221: error: (Each undeclared identifier is reported only once mailslot.c:221: error: for each function it appears in.) make[1]: *** [mailslot.o] Error 1 make[1]: Leaving directory `/var/tmp/portage/wine-0.9.10/work/wine-0.9.10/server' Suggested fix: Include stdint.h in server/mailslot.c (Patch not included due to the simplicity of code changes)

2 1

CVS EU Server
by Sven.Paschukat＠t-online.de March 23, 2006

March 23, 2006

Does anyone know what's wrong with the CVS server rhlx01.fht-esslingen.de in Europe? It's out of sync for a few days now. Sven

3 5

Regression 16 bit threading problem with Use explorer to handle desktop window commit
by paul March 23, 2006

March 23, 2006

If I start two 16 bit programs in wcmd or in the original case use wintask to start a Delphi 1 based GUI the second program does not get a new thread detach. As soon as the switch occurs back to the original task a page fault occurs. See bug for traces, more details and sample program. http://bugs.winehq.org/show_bug.cgi?id=4897 Paul Romanyszyn

1 0

Dev-C++ dogfooding
by Segin March 23, 2006

March 23, 2006

So far, everything works, except when you try to install a package that already is installed (of any version, so this also breaks upgrades via the package manager) Package manager starts up, presents you with a disloag (like aother package install) but when it goes to copy files, it open a dialog box with the error "Cannot Open AVI" with buttons which are clipped off (the window is a bit too small, push the farthest to the left to makei t resize and such.) It should be noted that a manual upgrade can be done by untarring the DevPak (it's a .tar.bz2) and copying the files manually, if the directory names aren't mangled. For some odd reason some files won't compile, yet others do, i can't explain it. Must be a MingW bug.

6 6

Another fun winelib project
by Steven Edwards March 23, 2006

March 23, 2006

It seems Microsoft has MS-FOSSed Mech Commander 2 under a shared source license http://www.microsoft.com/downloads/details.aspx?FamilyID=6d790cde-c3e5-46be… According to the appdb it does not run so maybe having the source could better help us find out what we are missing. -- Steven Edwards "There is one thing stronger than all the armies in the world, and that is an idea whose time has come." - Victor Hugo

2 1

Re: Regression in gdi
by Tony Lambregts March 23, 2006

March 23, 2006

Alexander N. Sørnes wrote: > Greetings, > > The patch mentioned below introduced a regression in gdi, so that the > part of the slidefields in Windows Media Player 9 left of the slider > were drawn as single-colour rectangles instead of a normal-looking part > of the slidefield. It can be unapplied cleanly from current CVS. See > bug 4868 (http://bugs.winehq.org/show_bug.cgi?id=4868) > > gdi: Better support for 1bpp dib sections. > http://source.winehq.org/git/?p=wine.git;a=commit;h=b55746ab62739fc1e4c0d4a… > > It's not just windows mediaplayer thats for sure. winrar also suffers from this regression. (dogfood) -- Tony Lambregts

1 0

Re: Store GL context in the TEB
by Tomas Carnecky March 23, 2006

March 23, 2006

Mike Hearn wrote: > Mike Hearn <mike(a)plan99.net> > Optimize thunks by storing GL context in the thread environment block > good job. this fixed the BadMatch error in World of Warcraft and also increased performance, from ~20fps to ~25fps. tom

3 4

Keyboard fix - patch available
by penna＠bb.com.br March 23, 2006

March 23, 2006

Hi all, Sorry for bothering you guys ..... but .... I saw a patch for bug #2400 (a very old bug) at http://bugs.winehq.org/show_bug.cgi?id=2400 Looking at the comments at the bugzilla tracking, the patch worked flawlessly for lots of keyboard types (croatian, german and russian), including mine: a brazilian abnt2 keyboard. The patch looks fine and very clean. The problem mentioned with remote X-windows access seems not to occur besides someone mention it. Anyway, at least the problem is gone with local access and, *if confirmed*, remains with remote access only. Maybe, just maybe, the patch could be applied for the next version of wine (0.9.11) .... ;-) Thank you very much for the very good piece of work: WINE (I use it since 1999). --------------------------------------------------------- Ulisses de Sousa Penna Analista Consultor - Banco do Brasil Fone: +55-61-3310-6320 Fax: +55-61-3310-6435 --------------------------------------------------------- P.S.: I am not subscribed at wine-devel list (just wine-users list)

1 0

IDE with Winemaker
by Ng, Billy H CIV NAVSEA March 23, 2006

March 23, 2006

Hi, I am a Windows Developer who is new to the Linux world. Is there any IDE (like Visio Studio) I can use with Winemaker? I am trying to port a Windows App to Linux Fedora. Thanks, Billy.

3 2

Wiki access
by Tom Spear (Dustin Booker, Dustin Navea) March 23, 2006

March 23, 2006

Could someone give me access to rename and delete pages in the wiki? I can't even rename or delete my own user page (!!) Thanks Tom

1 0

Re: advpack: Forward the remaining advpack.c ansi functions to their unicode counterparts
by Alexandre Julliard March 23, 2006

March 23, 2006

"James Hawkins" <truiken(a)gmail.com> writes: > +HRESULT WINAPI SetPerUserSecValuesA(PERUSERSECTIONA* pPerUser) > +{ > + PERUSERSECTIONW perUserW; > + > + if (!pPerUser) > + return E_INVALIDARG; > + > + TRACE("(%p)\n", pPerUser); > + > + /* lengths copied from PERUSERSECTION structure definition */ > + MultiByteToWideChar(CP_ACP, 0, pPerUser->szGUID, 59, perUserW.szGUID, 59); > + MultiByteToWideChar(CP_ACP, 0, pPerUser->szDispName, 128, perUserW.szDispName, 128); > + MultiByteToWideChar(CP_ACP, 0, pPerUser->szLocale, 10, perUserW.szLocale, 10); > + MultiByteToWideChar(CP_ACP, 0, pPerUser->szStub, MAX_PATH * 4, perUserW.szStub, MAX_PATH * 4); > + MultiByteToWideChar(CP_ACP, 0, pPerUser->szVersion, 32, perUserW.szVersion, 32); > + MultiByteToWideChar(CP_ACP, 0, pPerUser->szCompID, 128, perUserW.szCompID, 128); The source lengths are wrong, these are null-terminated strings. Also please use sizeof() instead of hardcoding the constants. -- Alexandre Julliard julliard(a)winehq.org

3 2

Re: winedbg: Fix stepping over
by Eric Pouech March 22, 2006

March 22, 2006

Petr Tesarik wrote: > Hi, > > this patch fixes a bug in winedbg, which caused it to continue > execution at a wrong address when stepping over a function call. > The trouble was that if there was no breakpoint found, the code at the > end of break_should_continue reversed the PC adjustment and later the > debugger resumed program execution AFTER the INT3 instruction, which > is clearly wrong. > Index: break.c > =================================================================== > RCS file: /home/wine/wine/programs/winedbg/break.c,v > retrieving revision 1.17 > diff -u -r1.17 break.c > --- break.c 27 Jan 2006 15:22:15 -0000 1.17 > +++ break.c 21 Mar 2006 10:36:07 -0000 > @@ -731,9 +731,13 @@ > if (code == EXCEPTION_BREAKPOINT) > addr->Offset += be_cpu->adjust_pc_for_break(&dbg_context, TRUE); > > + dbg_curr_thread->stopped_xpoint = find_xpoint(addr, be_xpoint_break); > dbg_curr_process->bp[0].enabled = FALSE; /* disable the step-over breakpoint */ > > - dbg_curr_thread->stopped_xpoint = find_xpoint(addr, be_xpoint_break); this part looks correct > + /* Unconditionally stop on the step-over breakpoint */ > + if (dbg_curr_thread->stopped_xpoint == 0) > + return FALSE; > + but this doesn't... you don't handle the count here... what you should do is only compute dbg_curr_thread->stopped_xpoint (and do the appropriate handling) when its current value is -1 (not found) A+ -- Eric Pouech

1 0

Easy applications to write automated tests for?
by Dan Kegel March 22, 2006

March 22, 2006

I noticed a probable regression in an app (kid pix 3) tonight. Which got me thinking: what are the easiest set of application regression tests we could throw together that go further than just verifying the installer runs? I'm thinking not just of things that could be automated with cxtest, but also stuff that might not fit into that framework. In the case of Kid Pix 3, a test that simply started the app, waited ten seconds, and made sure an error dialog didn't pop up would have detected the problem. ('Course, only a few people would be able to run that, as Kid Pix 3 is not a free download.) cxtest should be able to handle that. A bit more complex would be to launch WinZip, have it compress one file, and make sure the compressed file was as expected. It shouldn't be too hard to make cxtest handle that. And then on the somewhat ambitious end, now that Sun's JVM basically installs and runs trivial programs, we could run through the Java regression test suite at http://sourceware.org/mauve/ Heck, I might give that a shot myself. There's probably no reason to try to shove that into cxtest, is there? Or would it fit? And beyond the looking glass, one could imagine running the OpenOffice 2.0 automated test suite. (That's unlikely to work well at the moment, given that OpenOffice 2.0's open dialog locks up reliably.) That would probably flush out a few problems. - Dan -- Wine for Windows ISVs: http://kegel.com/wine/isv

4 11

Re: WIDL: Add missing ignore attribute. Check for inapplicable attributes in typedef.
by Alexandre Julliard March 22, 2006

March 22, 2006

Saveliy Tretiakov <saveliyt(a)mail.ru> writes: > ChangeLog: > Saveliy Tretiakov <saveliyt(a)mail.ru> > - Add missing ignore attribute. > - Check for inapplicable attributes in typedef. Changed attribute > parsing in a way Mike McCormack suggested. > - Register type alias in typelib, if ATTR_PUBLIC is set. I think checking values explicitly is preferable to encoding them in the syntax, since it gives you a chance to provide meaningful errors. If you rely on the parser all you'll get is a generic "syntax error" with no clue to what you did wrong. -- Alexandre Julliard julliard(a)winehq.org

2 1

Re: advpack: Forward AddDelBackupEntryA to its unicode counterpart
by Alexandre Julliard March 22, 2006

March 22, 2006

"James Hawkins" <truiken(a)gmail.com> writes: > + do > + { > + MultiByteToWideChar(CP_ACP, 0, ansi_list, len, ptr, len); > + > + ansi_list += len; > + ptr += len; > + > + len = lstrlenA(ansi_list) + 1; > + size += len * sizeof(WCHAR); > + list = HeapReAlloc(GetProcessHeap(), 0, list, size); > + } while (*ansi_list); You are assuming that A and W have the same length, that's wrong. Also it will be quite expensive to resize the list all the time. You should determine the full length of the list and then convert everything in one go. -- Alexandre Julliard julliard(a)winehq.org

1 0

linux mmap bug ? not filling in allocation gaps
by Geoff Brimhall March 22, 2006

March 22, 2006

I've been debugging why an n64 emulator ( Project64 ) gets "Failed to allocate memory" errors under wine 9.9.10 on linux kernel 2.6. The program allocates a large chunk of size 0x20100000, which succeeds, and then another large 0x10100000 chunk, which fails. WINEDEBUG=+relay,+virtual eventually brought me to look at the linux /proc/<pid>/maps to see if the address space was really fully allocated. Here's what I noticed: Address start - end: 00400000-0554000 - Project64.exe address 00554000-20000000 ---p large mem allocation *** <gap> *** 55555000-573e0000 - loaded shared libraries 573e0000-774e0000 ---p mem alloc of 0x20100000 7bf00000-7c003000 - wine-pthread 7c003000-7c1ee000 - [heap] 7ffe0000-fffe0000 ---p large mem allocation ffff7000-fffff000 - [stack] Wrote a test mmap binary to fill a process address space, it created a similar gap between the executable load address and the loaded shared libraries, eventually failing mmap calls when it hit the stack address space. When I updated the test app to explicitly mmap in the gap area via specifying the aligned start address, the mmap call succeeded. I can't be the first person to notice this. Are there any TODO solutions for this ? One linux-centric solution would be to look for gaps like this in /proc/<pid>/maps in the libs/wine/mmap.c and take advantage of this. Or maybe another would be to keep a table of address spaces returned in wine_anon_mmap as a guide, then validate with try_mmap_fixed as to where a next-allocation should occur ( would this solution work outside linux ) ? Thanks, geoff Brimhall __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com

3 2