September 2011 - Wine-devel

Re: [PATCH] cmd: Avoid rereading batch file for every call/goto executed (try 3)
by Dan Kegel 18 Sep '11

18 Sep '11

2011/9/18 Dan Kegel <dank(a)kegel.com>: > Don't let this one languish - it makes a huge difference for me! I should mention: with your latest additions to the cmd tests, time to run the cmd tests is starting to go off the cliff, and is now above four minutes... but with your efficiency patch, it only takes four seconds.

3 2

Unimplemented function?
by Luis Carlos Busquets Pérez 18 Sep '11

18 Sep '11

It seems that there is a regression in wine 1.3.28. Before bi-secting, I woul dlike to understand how that can happen. Apparently PlaChessV5.exe crashes due to: wine: Call from 0x68215ca8 to unimplemented function msvcrt.dll._snwprintf_s, aborting but when I look at the spec file, the function is defined: @ varargs _snwprintf_s(ptr long long wstr) MSVCRT__snwprintf_s and it exists in wcs.c How is it that them it results as undefined? Luis

2 2

Creating register Resource.
by Alistair Leslie-Hughes 18 Sep '11

18 Sep '11

Hi, I'm attempting to create a RES file from the xmlparser.idl file. After I modify the Makefile with IDL_R_SRCS = xmlparser.idl The .res file is created, but no register information is generated. Have I missed something? Best Regards Alistair Leslie-Hughes

1 0

Re: [PATCH 17/19] vbscript: Added support for exit do statement
by Octavian Voicu 18 Sep '11

18 Sep '11

On Fri, Sep 16, 2011 at 2:30 PM, Jacek Caban <jacek(a)codeweavers.com> wrote: > --- a/dlls/vbscript/compile.c > +++ b/dlls/vbscript/compile.c > static HRESULT compile_while_statement(compile_ctx_t *ctx, while_statement_t *stat) > { > - unsigned start_addr; > + unsigned start_addr, prev_label; ... > + if(stat->stat.type != STAT_WHILE) { > + prev_label = ctx->while_end_label; > + if((ctx->while_end_label = alloc_label(ctx)) == -1) > + return E_OUTOFMEMORY; > + } ... > + if(stat->stat.type != STAT_WHILE) { > + label_set_addr(ctx, ctx->while_end_label); > + ctx->while_end_label = prev_label; > + } Hello, This patch introduces a compile warning (gcc 4.5.2, both x86 and amd64): ../compile.c:489:26: warning: ‘prev_label’ may be used uninitialized in this function While I can see there's no actual problem, how can we shut the compiler's mouth? It may even be the case that if we just put a prev_label = 0 initializer, static analyzers will start complaining it's a useless assignment. Octavian

1 0

re: Where is the best place to report a fscanf bug found under wine-1.3.27?
by Dan Kegel 17 Sep '11

17 Sep '11

> Today I discovered .. that the scanf family of functions was > introducing float (32-bit floating-point) noise into double (64-bit > floating-point) results. I see it here, too: $ sudo apt-get install mingw32 $ i586-mingw32msvc-gcc test_fscanf.c $ echo "1.1e+01 1.1e+00 1.1e-01 1.1e-02 1.1e-03 1.1e-04" | ~/wine-git/wine a.exe 1.10000000000000000e+001 1.10000000000000009e+000 1.10000001639127737e-001 1.10000003278255491e-002 1.10000004917383261e-003 1.10000006556511075e-004 You can work around it with 'winetricks vcrun6', after which the test program outputs: 1.10000000000000000e+001 1.10000000000000010e+000 1.10000000000000000e-001 1.09999999999999990e-002 1.10000000000000010e-003 1.10000000000000000e-004 So the problem is in wine's msvcrt.dll. You could narrow it down further by printing out the raw byte strings for the numbers and then sprinking that printout through the path from when scanf is called to where printf returns. (Or, I suppose, step through it with a debugger instead.) But the best place to report the bug is http://bugs.winehq.org - Dan

2 2

Re: [po]updated korean resource(2011-09-16)
by Francois Gouget 17 Sep '11

17 Sep '11

> #: attrib.rc:47 > -#, fuzzy > msgid "" > "ATTRIB - Displays or changes file attributes.\n" > "\n" The reason why this one was fuzzy is that the indentation changed. However the translation's indentation has not be updated (which might possibly be ok for Korean), and it also seems to be pretty inconsistent (e.g. between the +/- lines and the R/A/etc ones. Also the '[drive:]' should probably really be indented. #: cmd.rc:288 -#, fuzzy msgid "" "CMD built-in commands are:\n" "ASSOC\t\tShow or modify file extension associations\n" The translation is missing the XCOPY line (it's hard to spot in the middle of all these lines). Maybe I should add translator comments when I mark a translation as fuzzy. This way the real translator would not have to puzzle over what caused the string to be marked as fuzzy. I have not been doing it because I worry they may be tricky to remove with some PO editors (but maybe my fears are unfounded). -- Francois Gouget <fgouget(a)free.fr> http://fgouget.free.fr/ Linux, WinNT, MS-DOS - also known as the Good, the Bad and the Ugly.

1 0

re: Severe startup latencies for Windows applications run under wine
by Dan Kegel 17 Sep '11

17 Sep '11

Try oprofile instead. ( A really old intro is at http://www.winehq.org/wwn/249#oprofile%20&%20Wine )

4 4

Proof of concept - The crashbot
by Bruno Jesus 17 Sep '11

17 Sep '11

Hi, I would like your opinion about a project I may start based on BuildBot (if the Dan/license allows). It's named crashbot because it's intended to automate testing of applications that crash in wine. The main goal is reduce the amount of time wasted on retesting bugzilla applications that authors abandoned. It's intented to be as easy as this: 1 - Add a new application by setting it's name, bug number and direct download url; 2 - Add requirements to run using winetricks; 3 - Set the test schedule (interval in days/weeks between repetitions); 4 - See the results in a web page which will automatically update; 5 - Auto send an email if the application starts successfully. The crashbot will run in a tmpfs to avoid disk burning and will ever run applications in a clean prefix. It would extract the downloaded program if necessary. Wine would be recompiled daily based on latest git. In the future I would like to add support to some sort of window scripting with keyboard and mouse input so it can be used not only to test startup crashs but any other kind of test (including the program installation before run). I'm not targeting it to test games for now because it would run in a very basic computer with an embedded video board. Thanks in advance and apologies if this already exists or was discussed previously. Best wishes, Bruno -- universe* god::bigbang (void); //and then it all began...

4 3

amd processors seem slow at running configure?
by Dan Kegel 17 Sep '11

17 Sep '11

My new AMD a8-3850 takes 77 seconds to run wine's configure script. (Using -C cuts the time by 40 seconds.) Austin's Phenom II X4 940 takes 61 seconds. My i7 takes 32 seconds... but even the lowly e7300 only takes 38 seconds. The same seems to hold for doing builds with ccache; AMD fares poorly there somehow compared to similar intel chips. Does anyone have an AMD-based system that can run configure in under 40 seconds?

5 6

re: Proof of concept - The crashbot
by Dan Kegel 16 Sep '11

16 Sep '11

Hi Bruno, it would of course be great. Check out http://code.google.com/p/winetricks/source/browse/trunk/tests/winetricks-te… for starters, we could just run that (or a subset of it) daily. It's fairly reliable, as I ran it manually every week for a while and weeded out a lot of flaky bits. When I get a chance, I'll try integrating that into buildbot. - Dan p.s. As Henri mentioned, it's been on people's minds for years. The graveyard is littered with previous attempts, e.g. http://web.archive.org/web/20070209153523/http://www.cxtest.org/ I think it's important to do the easy part first and not do anything fancy.

2 1