August 2004 - Wine-Devel - list.winehq.org

Re: Backtrace Dumps

by Robert Shearman

Alexandre Julliard wrote: >Andreas Mohr <andi(a)rhlx01.fht-esslingen.de> writes: > > > >>I'm sure every semi-involved Wine developer can imagine dozens of >>"reasons of the day" why winedbg doesn't launch properly on error again... >>Failure in wine exception handling code, failure to look up winedbg >>(both registry and disk), failure to pass winedbg cmdline parameters properly, >>failure to get winedbg started up properly, failure to get winedbg to >>parse the current modules and stack frame properly, ... >> >> > >Well, I can also imagine a lot of cases where the in-process backtrace >won't work right, the main one being that since the code will never be >used unless you want a backtrace, when you try to use it you'll >realize that it no longer works, or doesn't even compile anymore. If >we call the debugger instead, it will use the standard backtrace code >which is always compiled in, and regularly used by a lot of people. > > Do you want me to write a unit test that uses it then? Rob

Aug. 18, 2004

Re: OLEAUT32: implementation for IPersistPropertyBag_Load for OLEFont

by Robert Shearman

Alex Villacís Lasso wrote: > This implementation is enough to satisfy VB6 when loading font objects > from property bags in projects. > > Changelog: > * Added implementation for IPersistPropertyBag_Load on OLEFont Looks good, but a few comments below: ... >+ V_VT(&valueAttr) = VT_EMPTY; > > It is probably better practice to use VariantInit here. >+ iRes = IPropertyBag_Read(pPropBag, sAttrName, &valueAttr, pErrorLog); >+ switch (iRes) { > > >+ case S_OK: >+ if (V_VT(&valueAttr) == VT_BSTR) { >+ /* Type is BSTR, use for name directly */ >+ OLEFontImpl_put_Name(this, V_BSTR(&valueAttr)); >+ >+ /* FIXME: Leaking memory here? */ >+ } else { >+ /* Don't know how to handle this type */ >+ FIXME("Name has type %d, not implemented\n", V_VT(&valueAttr)); >+ iRes = E_NOTIMPL; >+ } > Perhaps use VariantChangeType here? >+ break; >+ case E_POINTER: >+ FIXME("\tiResult = [E_POINTER]\n"); >+ break; >+ case E_INVALIDARG: >+ FIXME("\tiResult = [undef]\n"); >+ iRes = S_OK; >+ break; >+ case E_FAIL: >+ FIXME("\tiResult = [E_FAIL]\n"); >+ break; >+ } > Catch the errors you need to (why is E_INVALIDARG ok?) and either move these common debug statements to the end of the function (they should also be ERR's, not FIXME's). Also, you other unexpected errors, which is bad. >+ } >+ >+ if (iRes == S_OK) { >+ CY iSize; >+ iSize.s.Hi = 0; >+ iSize.s.Lo = 0; >+ >+ V_VT(&valueAttr) = VT_EMPTY; >+ iRes = IPropertyBag_Read(pPropBag, sAttrSize, &valueAttr, pErrorLog); >+ switch (iRes) { >+ case S_OK: >+ switch (V_VT(&valueAttr)) { >+ case VT_I2: >+ iSize.s.Lo = V_I2(&valueAttr) * 10000UL; >+ break; >+ case VT_UI2: >+ iSize.s.Lo = V_UI2(&valueAttr) * 10000UL; >+ break; >+ case VT_I4: >+ iSize.s.Lo = V_I4(&valueAttr) * 10000UL; >+ break; >+ case VT_UI4: >+ iSize.s.Lo = V_UI4(&valueAttr) * 10000UL; >+ break; >+ case VT_R4: >+ iSize.s.Lo = (ULONG)(V_R4(&valueAttr) * 10000UL); >+ break; >+ case VT_R8: >+ iSize.s.Lo = (ULONG)(V_R8(&valueAttr) * 10000UL); >+ break; > > Again, perhaps use VariantChangeType here? >+ default: >+ /* Don't know how to handle this type */ >+ FIXME("Size type %d, not implemented\n", V_VT(&valueAttr)); >+ iRes = E_NOTIMPL; >+ break; >+ } >+ if (iRes != E_NOTIMPL) IFont_put_Size(this, iSize); >+ break; >+ case E_POINTER: >+ FIXME("\tiResult = [E_POINTER]\n"); >+ break; >+ case E_INVALIDARG: >+ FIXME("\tiResult = [undef]\n"); >+ iRes = S_OK; >+ break; >+ case E_FAIL: >+ FIXME("\tiResult = [E_FAIL]\n"); >+ break; >+ } > Same comment as before about errors. >+ } >+ if (iRes == S_OK) { >+ signed short iCharset = 0; >+ >+ V_VT(&valueAttr) = VT_EMPTY; >+ iRes = IPropertyBag_Read(pPropBag, sAttrCharset, &valueAttr, pErrorLog); >+ switch (iRes) { >+ case S_OK: >+ switch (V_VT(&valueAttr)) { >+ case VT_I2: >+ if (V_I2(&valueAttr) < 0) { >+ FIXME("Unexpected value for Font Charset (%hd), using 0\n", V_I2(&valueAttr)); >+ } else { >+ iCharset = V_I2(&valueAttr); >+ } >+ break; >+ case VT_UI2: >+ if (V_UI4(&valueAttr) > 32767L) { >+ FIXME("Unexpected value for Font Charset (%ld), using 0\n", V_UI4(&valueAttr)); >+ } else { >+ iCharset = V_UI2(&valueAttr); >+ } >+ break; >+ case VT_I4: >+ if (V_I4(&valueAttr) < 0 || V_I4(&valueAttr) > 32767L) { >+ FIXME("Unexpected value for Font Charset (%ld), using 0\n", V_I4(&valueAttr)); >+ } else { >+ iCharset = V_I4(&valueAttr); >+ } >+ break; >+ case VT_UI4: >+ if (V_UI4(&valueAttr) > 32767L) { >+ FIXME("Unexpected value for Font Charset (%ld), using 0\n", V_UI4(&valueAttr)); >+ } else { >+ iCharset = V_UI4(&valueAttr); >+ } >+ break; >+ default: >+ /* Don't know how to handle this type */ >+ FIXME("Charset type %d, not implemented\n", V_VT(&valueAttr)); >+ iRes = E_NOTIMPL; > > Use VariantChangeType here and you will get all the bounds checking for free. >+ } >+ if (iRes != E_NOTIMPL) IFont_put_Charset(this, iCharset); >+ break; >+ case E_POINTER: >+ FIXME("\tiResult = [E_POINTER]\n"); >+ break; >+ case E_INVALIDARG: >+ FIXME("\tiResult = [undef]\n"); >+ iRes = S_OK; >+ break; >+ case E_FAIL: >+ FIXME("\tiResult = [E_FAIL]\n"); >+ break; >+ } >+ } >+ >+ if (iRes == S_OK) { >+ signed short iWeight = FW_NORMAL; >+ >+ V_VT(&valueAttr) = VT_EMPTY; >+ iRes = IPropertyBag_Read(pPropBag, sAttrWeight, &valueAttr, pErrorLog); >+ switch (iRes) { >+ case S_OK: >+ switch (V_VT(&valueAttr)) { >+ case VT_I2: >+ if (V_I2(&valueAttr) < 0) { >+ FIXME("Unexpected value for Font Weight (%hd), using FW_NORMAL\n", V_I2(&valueAttr)); >+ } else { >+ iWeight = V_I2(&valueAttr); >+ } >+ break; >+ case VT_UI2: >+ if (V_UI4(&valueAttr) > 32767L) { >+ FIXME("Unexpected value for Font Weight (%ld), using FW_NORMAL\n", V_UI4(&valueAttr)); >+ } else { >+ iWeight = V_UI2(&valueAttr); >+ } >+ break; >+ case VT_I4: >+ if (V_I4(&valueAttr) < 0 || V_I4(&valueAttr) > 32767L) { >+ FIXME("Unexpected value for Font Weight (%ld), using FW_NORMAL\n", V_I4(&valueAttr)); >+ } else { >+ iWeight = V_I4(&valueAttr); >+ } >+ break; >+ case VT_UI4: >+ if (V_UI4(&valueAttr) > 32767L) { >+ FIXME("Unexpected value for Font Weight (%ld), using FW_NORMAL\n", V_UI4(&valueAttr)); >+ } else { >+ iWeight = V_UI4(&valueAttr); >+ } >+ break; >+ default: >+ /* Don't know how to handle this type */ >+ FIXME("Weight type %d, not implemented\n", V_VT(&valueAttr)); >+ iRes = E_NOTIMPL; >+ } > Again, use VariantChangeType. >+ if (iRes != E_NOTIMPL) IFont_put_Weight(this, iWeight); >+ break; >+ case E_POINTER: >+ FIXME("\tiResult = [E_POINTER]\n"); >+ break; >+ case E_INVALIDARG: >+ FIXME("\tiResult = [undef]\n"); >+ iRes = S_OK; >+ break; >+ case E_FAIL: >+ FIXME("\tiResult = [E_FAIL]\n"); >+ break; >+ } >+ } >+ >+ if (iRes == S_OK) { >+ BOOL bUnderline = FALSE; >+ >+ V_VT(&valueAttr) = VT_EMPTY; >+ iRes = IPropertyBag_Read(pPropBag, sAttrUnderline, &valueAttr, pErrorLog); >+ switch (iRes) { >+ case S_OK: >+ switch (V_VT(&valueAttr)) { >+ case VT_I2: >+ bUnderline = V_I2(&valueAttr) ? TRUE : FALSE; >+ break; >+ case VT_I4: >+ bUnderline = V_I4(&valueAttr) ? TRUE : FALSE; >+ break; >+ case VT_UI2: >+ bUnderline = V_UI2(&valueAttr) ? TRUE : FALSE; >+ break; >+ case VT_UI4: >+ bUnderline = V_UI4(&valueAttr) ? TRUE : FALSE; >+ break; >+ case VT_BOOL: >+ bUnderline = V_BOOL(&valueAttr); >+ break; >+ default: >+ /* Don't know how to handle this type */ >+ FIXME("Underline type %d, not implemented\n", V_VT(&valueAttr)); >+ iRes = E_NOTIMPL; >+ break; >+ } > Again, VariantChangeType. >+ if (iRes != E_NOTIMPL) IFont_put_Underline(this, bUnderline); >+ break; >+ case E_POINTER: >+ FIXME("\tiResult = [E_POINTER]\n"); >+ break; >+ case E_INVALIDARG: >+ FIXME("\tiResult = [undef]\n"); >+ iRes = S_OK; >+ break; >+ case E_FAIL: >+ FIXME("\tiResult = [E_FAIL]\n"); >+ break; >+ } >+ } >+ >+ if (iRes == S_OK) { >+ BOOL bItalic = FALSE; >+ >+ V_VT(&valueAttr) = VT_EMPTY; >+ iRes = IPropertyBag_Read(pPropBag, sAttrItalic, &valueAttr, pErrorLog); >+ switch (iRes) { >+ case S_OK: >+ switch (V_VT(&valueAttr)) { >+ case VT_I2: >+ bItalic = V_I2(&valueAttr) ? TRUE : FALSE; >+ break; >+ case VT_I4: >+ bItalic = V_I4(&valueAttr) ? TRUE : FALSE; >+ break; >+ case VT_UI2: >+ bItalic = V_UI2(&valueAttr) ? TRUE : FALSE; >+ break; >+ case VT_UI4: >+ bItalic = V_UI4(&valueAttr) ? TRUE : FALSE; >+ break; >+ case VT_BOOL: >+ bItalic = V_BOOL(&valueAttr); >+ break; >+ default: >+ /* Don't know how to handle this type */ >+ FIXME("Italic type %d, not implemented\n", V_VT(&valueAttr)); >+ iRes = E_NOTIMPL; >+ break; >+ } > Again, use VariantChangeType. >+ if (iRes != E_NOTIMPL) IFont_put_Italic(this, bItalic); >+ break; >+ case E_POINTER: >+ FIXME("\tiResult = [E_POINTER]\n"); >+ break; >+ case E_INVALIDARG: >+ FIXME("\tiResult = [undef]\n"); >+ iRes = S_OK; >+ break; >+ case E_FAIL: >+ FIXME("\tiResult = [E_FAIL]\n"); >+ break; >+ } >+ } >+ >+ if (iRes == S_OK) { >+ BOOL bStrikethrough = FALSE; >+ >+ V_VT(&valueAttr) = VT_EMPTY; >+ iRes = IPropertyBag_Read(pPropBag, sAttrStrikethrough, &valueAttr, pErrorLog); >+ switch (iRes) { >+ case S_OK: >+ switch (V_VT(&valueAttr)) { >+ case VT_I2: >+ bStrikethrough = V_I2(&valueAttr) ? TRUE : FALSE; >+ break; >+ case VT_I4: >+ bStrikethrough = V_I4(&valueAttr) ? TRUE : FALSE; >+ break; >+ case VT_UI2: >+ bStrikethrough = V_UI2(&valueAttr) ? TRUE : FALSE; >+ break; >+ case VT_UI4: >+ bStrikethrough = V_UI4(&valueAttr) ? TRUE : FALSE; >+ break; >+ case VT_BOOL: >+ bStrikethrough = V_BOOL(&valueAttr); >+ break; >+ default: >+ /* Don't know how to handle this type */ >+ FIXME("Strikethrough type %d, not implemented\n", V_VT(&valueAttr)); >+ iRes = E_NOTIMPL; >+ break; >+ } > > Again, use VariantChangeType. >+ if (iRes != E_NOTIMPL) IFont_put_Strikethrough(this, bStrikethrough); >+ break; >+ case E_POINTER: >+ FIXME("\tiResult = [E_POINTER]\n"); >+ break; >+ case E_INVALIDARG: >+ FIXME("\tiResult = [undef]\n"); >+ iRes = S_OK; >+ break; >+ case E_FAIL: >+ FIXME("\tiResult = [E_FAIL]\n"); >+ break; >+ } >+ } >+ return iRes; > } > > static HRESULT WINAPI OLEFontImpl_IPersistPropertyBag_Save( > > Thanks, Rob

Aug. 18, 2004

Re: Backtrace Dumps

by Alexandre Julliard

Andreas Mohr <andi(a)rhlx01.fht-esslingen.de> writes: > I'm sure every semi-involved Wine developer can imagine dozens of > "reasons of the day" why winedbg doesn't launch properly on error again... > Failure in wine exception handling code, failure to look up winedbg > (both registry and disk), failure to pass winedbg cmdline parameters properly, > failure to get winedbg started up properly, failure to get winedbg to > parse the current modules and stack frame properly, ... Well, I can also imagine a lot of cases where the in-process backtrace won't work right, the main one being that since the code will never be used unless you want a backtrace, when you try to use it you'll realize that it no longer works, or doesn't even compile anymore. If we call the debugger instead, it will use the standard backtrace code which is always compiled in, and regularly used by a lot of people. -- Alexandre Julliard julliard(a)winehq.org

Aug. 18, 2004

Re: Some more hardcoded c:\windows paths

by Juan Lang

Vincent wrote: > Any idea why CSIDL_PROGRAMS was HKEY_DISALLOWED? That, apparently, was my mistake. I introduced it here: http://www.winehq.org/hypermail/wine-patches/2004/04/0106.html Thanks for fixing it. --Juan __________________________________ Do you Yahoo!? New and Improved Yahoo! Mail - Send 10MB messages! http://promotions.yahoo.com/new_mail

Aug. 18, 2004

Re: WineSetupTk...

by Ivan Leo Puoti

> I doubt there's a one maintainer Nobody maintains it because it's obsolete and isn't used with current versions of wine. Ivan.

Aug. 18, 2004

Re: winetest can't parse subtest for msvcrt(d)

by Saulius Krasuckas

On Wed, 18 Aug 2004, Ferenc Wagner wrote: > Saulius Krasuckas <saulius2(a)ar.fi.lt> writes: > > I think that's due to some additional \r\n-s in the output > > of "msvcrt_test.exe.so". > > Actually, not quite. The problem is that msvcrt's output > functions aren't redirected by libc's filehandle operations, > so winetest doesn't get the output (it gets printed to the > screen instead). Oh.. :-/ > Yeah, it's annoying, but doesn't hurt the meant functionality of > winetest, which shouldn't be run under Wine for anything but testing > purposes. The native (Windows) run doesn't show this problem, since > there we use msvcrt functions exclusively. Still it would be nice to > fix, but the proper way would be to fix Wine, and that's not entirely > trivial as Alexandre commented. Can you provide a link to the post or maybe some keywords to find it using google? Thanks.

Aug. 18, 2004

Re: winetest can't parse subtest for msvcrt(d)

by Ferenc Wagner

Saulius Krasuckas <saulius2(a)ar.fi.lt> writes: > On Wed, 18 Aug 2004, Ferenc Wagner wrote: >> Saulius Krasuckas <saulius2(a)ar.fi.lt> writes: >>> I think that's due to some additional \r\n-s in the output >>> of "msvcrt_test.exe.so". >> >> Actually, not quite. The problem is that msvcrt's output >> functions aren't redirected by libc's filehandle operations, >> so winetest doesn't get the output (it gets printed to the >> screen instead). > > Oh.. :-/ > >> Yeah, it's annoying, but doesn't hurt the meant functionality of >> winetest, which shouldn't be run under Wine for anything but testing >> purposes. The native (Windows) run doesn't show this problem, since >> there we use msvcrt functions exclusively. Still it would be nice to >> fix, but the proper way would be to fix Wine, and that's not entirely >> trivial as Alexandre commented. > > Can you provide a link to the post or maybe some keywords to find it using > google? Thanks. Try this for example: http://search.gmane.org/search.php?query=redirect+winetest&email=&group=gma… If you can't help feeling like hacking on winetest, I can send you my todo list, too... :) Or maybe I should put it into the source so that innocent bycomers can see it... -- Cheers, Feri.

Aug. 18, 2004

Re: WineSetupTk...

by michael＠cherryblossom.homelinux.com

On Wed, Aug 18, 2004 at 09:20:17AM +0100, Mike Hearn wrote: > >I didn't realize that WineSetupTk was depreciated when I wrote. I just > >want to play StarCraft and Diablo and was finding the documentation a bit > >hard to follow. (I couldn't find my .wine.conf file and got stuck, that's > >why I thought I needed WineSetupTk.) I will poke at it a bit myself and if > >I can figure out how it works, maybe then I can make a suggestion or two > >for clearing up the documentation. (Who maintains the documentation, BTW?) I doubt there's a one maintainer - if there was, it'd be listed with the documentation in credits or something, but maybe u can submit the patches somehow... maybe try reading sf policies on submitting stuff to a project? > > OK, what specifically were you trying to configure. Our aim (obviously) > is that StarCraft and such should work out of the box with no > configuration needed. > It's worth noting that an installation process that works with release A might not work on release B because there are improvements in (or not in) release B that require a different or appendment to the process (than) in release A, regardless of which is older/newer.

Aug. 18, 2004

winetest can't parse subtest for msvcrt(d)

by Saulius Krasuckas

On linuxe I get this: [s2(a)katleriai s2]$ wine /usr/local/lib/wine/winetest.exe.so Usage: /tmp/wctW5ad4P/msvcrt_test.exe.so test_name Valid test names: cpp environ file headers heap scanf string time ..while Wine Test Shell pops up a message box with the error text: "Can't parse subtests output of msvcrt". I think that's due to some additional \r\n-s in the output of "msvcrt_test.exe.so". Here goes my observation and comparision to the case of "mlang_test.exe.so" (which works smoothly with the winetest): [s2(a)katleriai tests]$ wine msvcrt_test.exe.so | head -n 1 | od -t x1 0000000 55 73 61 67 65 3a 20 6d 73 76 63 72 74 5f 74 65 0000020 73 74 2e 65 78 65 2e 73 6f 20 74 65 73 74 5f 6e 0000040 61 6d 65 0d 0a 0000045 [s2(a)katleriai tests]$ wine mlang_test.exe.so | head -n 1 | od -t x1 0000000 55 73 61 67 65 3a 20 6d 6c 61 6e 67 5f 74 65 73 0000020 74 2e 65 78 65 2e 73 6f 20 74 65 73 74 5f 6e 61 0000040 6d 65 0a 0000043 As the "Valid test names:" string is printed by calling fprintf(), I suppose some differences arise due to incorrect linkage or such: [s2(a)katleriai tests]$ nm msvcrt_test.exe.so | grep fprintf 0001e100 D fprintf 0001e180 D vfprintf [s2(a)katleriai tests]$ nm mlang_test.exe.so | grep fprintf U fprintf@@GLIBC_2.0 U vfprintf@@GLIBC_2.0 It seems we realy are using msvcrt version of the fprintf(), which works in true-M$-way and adds an additional \r here. Next goes last lines from the build process: [s2(a)katleriai tests]$ (make clean && make) | tail -n 1 ../../../tools/winegcc/winegcc -B../../../tools/winebuild -mconsole cpp.o environ.o file.o headers.o heap.o scanf.o string.o time.o testlist.o -o msvcrt_test.exe.so -L../../../libs/port -lwine_port -L../../../dlls -L../../../libs -lmsvcrt [s2(a)katleriai tests]$ (make clean && make) | tail -n 1 ../../../tools/winegcc/winegcc -B../../../tools/winebuild -mconsole mlang.o testlist.o -o mlang_test.exe.so -L../../../libs/port -lwine_port -L../../../dlls -L../../../libs -lole32 -lgdi32 -lkernel32 -lntdll -luuid Can we fix it? I have no idea on how to link it right with the GLIBC. Of course, I am not sure this is the core of the problem. It would be nice to have winetest running in linux without such two hits (one for the msvcrt and another one for the msvcrtd). Or is this just my linux box configuration flawed?

Aug. 18, 2004

Re: winetest can't parse subtest for msvcrt(d)

by Ferenc Wagner

Saulius Krasuckas <saulius2(a)ar.fi.lt> writes: > I think that's due to some additional \r\n-s in the output > of "msvcrt_test.exe.so". Actually, not quite. The problem is that msvcrt's output functions aren't redirected by libc's filehandle operations, so winetest doesn't get the output (it gets printed to the screen instead). Yeah, it's annoying, but doesn't hurt the meant functionality of winetest, which shouldn't be run under Wine for anything but testing purposes. The native (Windows) run doesn't show this problem, since there we use msvcrt functions exclusively. Still it would be nice to fix, but the proper way would be to fix Wine, and that's not entirely trivial as Alexandre commented. -- Cheers, Feri.

Aug. 18, 2004