April 2009 - Wine-devel - list.winehq.org

DIB Engine - New approach
by Massimo Del Fedele 21 Apr '09

21 Apr '09

The approach taken so far consisted in having 2 device pointers inside GDI32, one for dib engine and the other for normal display driver. This way had the disadvantage of having to keep in sync the DC with the right driver depending on selected bitmap, which lead to many small changes spread along all gdi code; going deeper with development this approach showed many limits and problems. So I decided to start again from scratch with a completely different approach which is giving good results and is quite less invasive on gdi32 code. Instead of doing : -- DIB ENGINE / GDI32 --- \ -- X11 DRIVER I took this approach : GDI32 -- DIB ENGINE -- X11 DRIVER The (X11) display driver is loaded from inside the engine, which replaces it from gdi32 point of view. The changes in gdi32 are *very* limited, just in driver.c code, making it to load (if desired) the engine instead of normal x11 driver. No other changes needed. I added as usual the code allowing to enable/disable the engine on request by registry and/or environment variable. If the engine is not present or disabled, the driver loader falls back to usual behaviour. The Engine then loads the X11 driver in init phase and acts as a gate depending on selected BMP, forwarding to X11 driver all requests for DDB bitmaps and doing the job itself for DIB ones. This approach showed many advantages, and I' almost ready converting all old code to it. By now I'm posting here 3 patches showing the approach; the posted driver is a simple pass-throu one, so it just forwards all calls to X11 driver. The last patch of the series shows the forking behaviour of DIB/DDB processing from inside the engine, but still forwarding all to X11. I'd like some comments on the approach taken; on next days I'll post a more complete code with most of the engine implemented. Ciao Max

3 5

Re: [1/5]msctf/test: setup test framework for ITfThreadMgrEventSink
by Alexandre Julliard 21 Apr '09

21 Apr '09

Aric Stewart <aric(a)codeweavers.com> writes: > +static HRESULT WINAPI ThreadMgrEventSink_OnInitDocumentMgr(ITfThreadMgrEventSink *iface, > +ITfDocumentMgr *pdim) > +{ > + trace("%s\n",__FUNCTION__); > + return S_OK; > +} trace() already includes the file and line, and __FUNCTION__ is not portable, please avoid it in tests. -- Alexandre Julliard julliard(a)winehq.org

1 0

Wine patch code review or what to do if patches get stuck?
by Igor Tarasov 21 Apr '09

21 Apr '09

Hi! The reason of why I write this to you is simple: my patches have stuck again and I don't know what to do. But since this happening not with me only and not this happens regularly, I have thought about looking at the problem a bit wider. Is there any code review system for patches that are sent to wine-patches? Or just some developers monitor wine-patches and say something if they have something to say? Or, how does Alexander make decisions regarding committing patches? I've been thinking about some kind of web-based system, like one used at http://codereview.chromium.org/ It would be nice, if there was sime kind of bot, that would grab all patches from wine-patches, sort them, make some initial analysis (maybe, run local test git repository, make various tests, check that threre are no conflicts, check code style (at least whitespace changes) etc). Then it could set a flag and report back to the sender. For instance it would flag "tests fail" and send details back to sender. Or "whitespace change detected" and send back the quoted diff parts. Then, somebody of developers might review it and put his mark/comments that would go to wine-devel (if necessary), or to sender directly. There could be simple templates for test reviews for common problems, and tags, such as "sure?" that means, that more explanations are needed why that implementation is correct. "too hackish" that means that style or code itself is of bad quality. "whitespace" whitespacechange detected. "tests required" tests are really required for this patch to go in. "okay", means that reviewer recommends the patch for commitment. "wrong", means that patch is totally wrong ..etc Some of these tags mean that report goes directly to developer, and patch is marked as obsolete, others make system send some feedback to wine-dev. It would take just few clicks for reviewer to get it done. This would provide good feedback, also this would help the one who makes the decisions to see that others have already checked that patch and what they think about that. Sometimes "no feedback" means that you just dont know - is there something wrong? Or maybe your patch wos not understood correctly? Or there are doubts that what you did is right? Or what? It would be better to have feedback like "are you serious?" than nothing at all :) P.S: the patches in question are http://www.winehq.org/pipermail/wine-patches/2009-April/072009.html and http://www.winehq.org/pipermail/wine-patches/2009-April/072013.html . They both are soo simple, that I even don't know how to ask of what's wrong with them :) -- Igor

5 6

Re: [1/2](try 2)msctf: helper functions for generating DWORD cookies
by Alexandre Julliard 21 Apr '09

21 Apr '09

Aric Stewart <aric(a)codeweavers.com> writes: > + else if (id_last == array_size) > + { > + CookieInternal **new_cookies = NULL; > + new_cookies = HeapReAlloc(GetProcessHeap(),0,cookies,sizeof(CookieInternal*) * (array_size * 2)); > + if (!new_cookies) > + { > + array_size = 0; > + ERR("Out of memory, Unable to realloc cookies array\n"); > + return 0x0; > + } > + cookies = new_cookies; > + array_size *= 2; > + } > + > + /* try to reuse IDs if possible */ > + for (i = 0; i < id_last; i++) > + if (cookies[i] == NULL) break; It would be smarter to reuse the ID before growing the array. Also you could mark free entries with a 0 id or something like that, and avoid a level of indirection (and some memory leaks...) -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: [kernel32] add unittests for registry keys of serial ports
by Alexandre Julliard 21 Apr '09

21 Apr '09

Stefan Leichter <Stefan.Leichter(a)camline.com> writes: > + RegOpenKeyA(HKEY_LOCAL_MACHINE, "Hardware\\Description\\System", &desc_sys); > + RegOpenKeyA(HKEY_LOCAL_MACHINE, "Hardware\\Devicemap\\Serialcomm", &sercom); > + RegOpenKeyA(HKEY_LOCAL_MACHINE, "Software\\Microsoft\\Windows NT\\CurrentVersion\\Ports", &sw_ports); You need to check for failure and handle it properly. I doubt you'll be able to open these keys on Win9x for instance. > + result = RegOpenKeyA(HKEY_LOCAL_MACHINE, "System\\Select", &sub_key1); > + if(ERROR_SUCCESS == result) { > + size = sizeof(count1); > + result = RegQueryValueExA( sub_key1, "Current", NULL, NULL, (LPBYTE) &count1, &size); > + RegCloseKey(sub_key1); > + } > + if(ERROR_SUCCESS == result) { > + sprintf(sub_key1_name, "System\\ControlSet%03d\\Enum\\ACPI", count1); > + RegOpenKeyA(HKEY_LOCAL_MACHINE, sub_key1_name, &acpi); > + } Do you really have an app that accesses ControlSetxxx directly? CurrentControlSet is a lot easier to work with. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: Possible regression today's git wine-1.1.19-153-g55ad3fd
by Susan Cragin 21 Apr '09

21 Apr '09

>>>>>Ohhh... got a nice crash after a re-boot. >>>> >>>> Never mind that. This looks more promising. At least, to me it does. The actctx error I get all the time, but I've never seen the other, and it happend right as the system froze up. >>>> >>>> fixme:actctx:parse_depend_manifests Could not find dependent assembly L"Microsoft.VC80.MFCLOC" (8.0.50608.0) >>>> err:wave:wodPlayer_WriteMaxFrags Error in writing wavehdr. Reason: Broken pipe >>>> err:wave:wodPlayer_WriteMaxFrags Error in writing wavehdr. Reason: Broken pipe >>>> err:wave:wodPlayer_WriteMaxFrags Error in writing wavehdr. Reason: Broken pipe >>>> err:wave:wodPlayer_WriteMaxFrags Error in writing wavehdr. Reason: Success >>> >>> Interesting. You are running Ubuntu studio 8.10 without pulseaudio, right? >>> >>> Have you updated the system between version runs (that may have >>> reintroduced pulseaudio, for example)? >>> >>> Are you running any other sound applications (or the sound that is >>> played after boot on Ubuntu)? >>> >>> Could you perform a git bisect to identify the guilty change? >> >>Doing some investigation, I have found >>(http://alsa.opensrc.org/index.php/HowTo_Asynchronous_Playback) >>""" >>That is basically all there is to the callback. Note however, that >>this example does not include error checking, which is a good idea of >>course. Occasionally, you may have to restore from a buffer under-run, >>in case the computer was caught up in some other process and didn't >>get to calling your callback. Whenever this happens, the first >>operation you perform on the device, in this case >>snd_pcm_avail_update, will return -EPIPE (a 'Broken Pipe' error). You >>can restore from this by calling the snd_pcm_prepare function again. >>""" >> >>This may be what you are experiencing here. I'd need to look into this >>further, but to be honest, I'm a little out of depth with the Windows >>and Wine sound interfaces. >> >>- Reece > >Reece: >I purged pulseaudio and checked that it had not been re-installed. I run Ubuntu 9.04 with all updates. I also use a test version of alsa-driver since my sound card is difficult to recognize but so far have had only great results -- no impact on wine at all. >The problem I am experiencing manifests itself in the program freezing... could be the sound, could also be the buttons doing something odd. One button's on-off toggle switches suddenly with no input on my part. >Regression test is below. >Susan > >b958dfdc4769515c0ef56c401430dd3e3980aabf is first bad commit >commit b958dfdc4769515c0ef56c401430dd3e3980aabf >Author: Huw Davies <huw(a)codeweavers.com> >Date: Thu Apr 16 11:51:45 2009 +0100 > > ole32: Change helper function to return the matching entry. > >:040000 040000 2964979dee4f353e1d565f7c20d8fdcc534f006d 8744cfee2e33db4799f58f102825103f565e18b9 M dlls I filed bug #18133, which has more detail and a screenshot. Susan

1 0

Re: [PATCH 5/5] user32/tests: Fix tests on win9x.
by Paul Vriens 21 Apr '09

21 Apr '09

Huw Davies wrote: > --- > dlls/user32/tests/clipboard.c | 7 +++++-- > 1 files changed, 5 insertions(+), 2 deletions(-) > > > ------------------------------------------------------------------------ > > Was just about to sent a patch along those lines. Isn't it better to mark that win9x (and winME as well btw) one as broken()? -- Cheers, Paul.

2 4

Re: Article on wine development strategy
by Susan Cragin 21 Apr '09

21 Apr '09

I would like to put in my two cents for making Dragon NaturallySpeaking 10 work at about "50%." DNS 10 is a terrible program in a lot of ways. It's interface is over-engineered, clumsy, unintuitive, and packed with features that the average user never even looks at. People who don't know DNS say, "I'd love to dictate into word, format by voice command, etc." People who know DNS well, say "just give me accuracy and speed, and forget everything else." Those people usually tend to be the natural DNS constituency: either disabled or want to churn out a tremendous amount of text without typing. I think my point here is that happiness depends not only on % but on the type of program. Not all programs are created equal. Not all features are a good idea. Susan Cragin

8 9

Maybe already known stuff, but.....
by Massimo Del Fedele 21 Apr '09

21 Apr '09

Today I found this : #include "execinfo.h" void BT(void) { void *array[50]; size_t size; char **strings; size_t i; size = backtrace (array, 50); strings = backtrace_symbols (array, size); printf ("Obtained %zd stack frames.\n", size); for (i = 0; i < size; i++) printf ("%s\n", strings[i]); free (strings); } That allow to dump a backtrace (here max 50 frames) from inside a running program. Quite useful if app is non debuggable and don't run with +relay on because of copy protection. Ciao Max

2 2

NetBSD compile error
by Austin English 21 Apr '09

21 Apr '09

I just installed NetBSD 5.0RC4 on a spare hard drive, and tried compiling Wine. I hadn't seen this error before, but I'm not sure if the regression is on NetBSD or Wine's end. Google doesn't show too much for this error, other than that apparently NetBSD's libexec requires execname to not be NULL: case 2: /* ORIGIN */ if (execname == NULL) xerr(1, "execname not specified in AUX vector"); if ((ep = strrchr(p = execname, '/')) == NULL) xerr(1, "bad execname `%s' in AUX vector", execname); break; http://ftp.lahtermaher.org/pub/os/netbsd/cvs/src/libexec/ld.elf_so/expand.c… austin987$ make LD_LIBRARY_PATH="../../libs/wine:$LD_LIBRARY_PATH" ../../tools/wrc/wrc --nostdinc -I. -I. -I../../include -I../../include -D__WINESRC__ -D_ADVAPI32_ -foversion.res version.rc execname not specified in AUX vector: No such file or directory *** Error code 1 Can anyone shed some light on this error? -- -Austin

1 0