January 2013 - Wine-Devel - list.winehq.org

Re: riched20: ME_StrDup() is no longer used so remove it.
by Huw Davies Jan. 29, 2013

Jan. 29, 2013

On Tue, Jan 29, 2013 at 07:05:13AM +0100, Francois Gouget wrote: > --- > dlls/riched20/editor.h | 1 - > dlls/riched20/string.c | 5 ----- > 2 files changed, 6 deletions(-) I'm about to send in a patch that uses this again, so please hold off on this one. Huw.

1 0

Re: winemenubuilder: Try to wait for creating icon files (try 2).
by Sergey Guralnik Jan. 29, 2013

Jan. 29, 2013

On 2013-01-23 16:34, Sergey Guralnik wrote: > It supersedes next patch: > > From 4ca51cf16c5be2bcf34dddb7f33fe043c1d777c4 Mon Sep 17 00:00:00 > 2001 > From: Sergey Guralnik <serhio(a)etersoft.ru> > Date: Tue, 22 Jan 2013 15:58:20 +0200 > Subject: winemenubuilder: Use only existing file for icon creating. > > The logic is the same as before - wait for parent process termination > if > required icon file doesn't exist yet. But now used easier way to do > this. Is something wrong with this patch? It's in queue about a week without any feedback. -- Sergey

1 0

Re: [PATCH] winebuild: Use Clang to assemble if found. (try 5)
by Alexandre Julliard Jan. 28, 2013

Jan. 28, 2013

Charles Davis <cdavis5x(a)gmail.com> writes: > Try 5: Don't use strstr(3). > > The configure script invokes the C compiler to figure out if .cfi > pseudo-ops are supported, on the assumption that it will invoke the > system assembler. Winebuild, on the other hand, invokes the system > assembler directly. This caused a problem on Mac OS hosts when compiling > with Clang: configure would detect that .cfi ops were supported, because > Clang's integrated assembler supports them; however, when winebuild then > emitted assembly with .cfi ops in them, the system assembler would then > choke. This patch makes winebuild invoke Clang's assembler if it is > found in the path. It's causing warnings on my Mac: clang: warning: /usr/bin/clang: 'linker' input unused when '-c' is present clang: warning: /usr/bin/clang: 'linker' input unused when '-c' is present -- Alexandre Julliard julliard(a)winehq.org

1 0

Why 0.1.0.0 as local socket address instead of 127.0.0.1?
by Francois Gouget Jan. 27, 2013

Jan. 27, 2013

The ws2_32:sock fails on some of my machines with the following error: sock.c:4794: Test failed: Local socket address is different 0.1.0.0 != 127.0.0.1 Does anyone know why we GetAcceptExSockaddrs() would return 0.1.0.0 instead of 127.0.0.1? Looking at the Windows 7 winetest results it does not appear to be correlated with either the SP1 or IE9. For instance got an error on Windows 7 + SP1 + IE9 (32 and 64-bit): http://test.winehq.org/data/b233d2054e64e216a5b3868f301bfc934b019b4b/win7_f… http://test.winehq.org/data/b233d2054e64e216a5b3868f301bfc934b019b4b/win7_f… But not here (Windows 7 + SP1 + IE9 64-bit): http://test.winehq.org/data/83846cd131aac33228bf08aefae0bf0ea93d4273/win7_f… I also got it failing at least once on a Windows 2000 SP4 VM: http://test.winehq.org/data/b233d2054e64e216a5b3868f301bfc934b019b4b/2000_f… But never saw it fail on my Windows XP or 2003 VMs. -- Francois Gouget <fgouget(a)codeweavers.com>

2 1

FOSDEM logistics
by Jeremy White Jan. 27, 2013

Jan. 27, 2013

Hi folks, Alright, this is it - we're going to try to get together at FOSDEM starting this Friday. If you are planning on being there, you should make sure to subscribe to the wineconf mailing list: http://www.winehq.org/mailman/listinfo/wineconf All further logistics emails and discussions will take place there. So if you're not subscribed, you can't complain we didn't try to include you in the planning. We plan on informally assembling in the lobby / bar of the Best Western Hotel Royal Centre (http://www.royalcentre.be/) on Friday, most likely starting in late afternoon. I don't have any plans for Saturday; I'll likely go harangue the Xorg guys, and maybe catch the key note. We may try to informally gather as well. However, there is no plan for Saturday dinner (sorry, no sponsored dinner by CodeWeavers this year). I have no idea how FOSDEM usually works on Saturday night, so I figured we'd just play it be year. Advice is welcome. Our devroom starts Sunday at 9:00 am. We're in building K, which you can see on this map: http://www.ulb.ac.be/campus/solbosch/plan-K.html Main FOSDEM entrance / registration is in building H on that map. One bit of unsolicited advice - when I naively copy/paste the address on the FOSDEM page into Google maps for directions, I end up with the wrong location. If you're at the Best Western, it looks like Tram 94 goes to the Legrand stop in 16 minutes, and then a 10 minute walk carries you the rest of the way there. Looking forward to seeing you all on Friday! Cheers, Jeremy

1 0

Re: wined3d: Rebind texture before checking for its content in check_fbo_compat().
by Stefan Dösinger Jan. 27, 2013

Jan. 27, 2013

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Am 2013-01-26 22:32, schrieb Matteo Bruni: > + > gl_info->fbo_ops.glFramebufferTexture2D(GL_FRAMEBUFFER, > GL_COLOR_ATTACHMENT0, GL_TEXTURE_2D, tex, 0); Do you really have to re-attach it to the FBO? Or does glBindTexture(GL_TEXTURE_2D, tex) work as well? -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.19 (GNU/Linux) Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/ iQIcBAEBAgAGBQJRBFTlAAoJEN0/YqbEcdMw7bcQAIfUM81xBr3zIpcEYek3gMso jXPq49+8Dp8Y3xiJGU/nFD8yLH9Y24SXpyageueCMWZR2jnEQcXBL0l76MmL5ZYK Y/L1ytRwrOvdXP07pJ8p3ZXxmA6hl82fuNxMvYhlSYP8PVCoobhaH7MwGp9WPpGF Ij20eOOR3GQnDjb6M2qHHfP9s+mNREu5NrvZ76NCf/lURXLJyDECuMz+6TbSl4CM qv0xCKdtqxm13gTZio3ftKc3QZZUSstiYjUYQmFh5NXQe5oIr9j2/iueU2K0zA/M CqswN4divbFFbh7QwCIDaubwfGidRfiwIG3lCCiNpfIvnowQouELIoFKjIipQ1AL ihIIS7iyekasAxkErVZ3E8O8IJCKYbTC+tAXBfK3niKmSRfdUB9cmfo5nVZw4Xjb 3/Zz84WnjMlIznyBGsKyKX+VLKlqzXY5zCyboHuEWPxTXdS0FAbwHzz06Kv0zIqc Pe4X6dJOUz6EPnx2gkbtFygGFbAa8haWhiIcY01H1bMSa7NnZPU3QgQXimrvXaKK SbQofZfhbNSVv0dL1kxqUYeLPb5hKXqMRHO7yi/GeyejTie9onzjxrjt4zn4rxe3 9IEZ0ymUYs499ahUnW/DAe0yFJvX+EXiPIiWwLBGVpx/Dzduk0WADOLBaN/YwrEz 1gxkuQjF2tbVSTkP41yr =xi8z -----END PGP SIGNATURE-----

3 3

Implement D3DXSHEvalConeLight
by Nozomi Kodama Jan. 26, 2013

Jan. 26, 2013

Hello what is the problem with the patch http://source.winehq.org/patches/data/93498 Nozomi

1 0

Re: [PATCH] gdi32: Adjust a fixed-pitch full-width character has double advance of a half-width character. (try 5)
by Alexandre Julliard Jan. 25, 2013

Jan. 25, 2013

Akihiro Sagawa <sagawa.aki(a)gmail.com> writes: > Hi, > > This patch enhances full-width character glyph advance when we choose > fixed-pitch font. > > Windows 2000 (or XP for Japanese) and later adjust fixed-pitch > full-width character glyph advances so that two half-width characters' > advance equals to one full-width character's advance. Because CJK > fixed-pitch font, such as MS Gothic, is designed as full-width character > (e.g. Kanji) has a double advance of half-width character (e.g. ASCII) > and many developpers expect that. > > Please refer Microsoft document KB 417434 [1] and white paper [2] for > details. > > [1] ... http://support.microsoft.com/kb/417434/ja (written in Japanese) > [2] ... http://msdn.microsoft.com/ja-jp/library/cc979401.aspx (written in Japanese) > > try 5: > - Add patch description above. > - Fix avgAdvance's rounding method. (as in get_outline_text_metrics) > - Refine a commit message, "Ensure" to "Adjust" and "double width" to > "double advance". > > try 4: > - Fix full-width character detection that causes test failure. > > try 3: > - Improve full-width character detection. > - Support fake italic transformation. > - Choose non-linked fonts in the test case. > - Add more trace messages. > > try 2: > - Add rotated transformation tests. > > Regards, > Akihiro Sagawa > > --- > dlls/gdi32/freetype.c | 37 ++++++++++++++++++-- > dlls/gdi32/tests/font.c | 87 +++++++++++++++++++++++++++++++++++++++++++++++ > 2 files changed, 121 insertions(+), 3 deletions(-) This is still failing here: ../../../tools/runtest -q -P wine -M gdi32.dll -T ../../.. -p gdi32_test.exe.so font.c && touch font.ok font.c:2549: Tests skipped: Font Droid Arabic Naskh doesn't contain 'x', skipping the test font.c:2549: Tests skipped: Font Droid Sans Armenian doesn't contain 'x', skipping the test font.c:2549: Tests skipped: Font Droid Sans Ethiopic doesn't contain 'x', skipping the test font.c:2549: Tests skipped: Font Droid Sans Georgian doesn't contain 'x', skipping the test font.c:2549: Tests skipped: Font Droid Sans Hebrew doesn't contain 'x', skipping the test font.c:2549: Tests skipped: Font Droid Sans Hebrew doesn't contain 'x', skipping the test font.c:3008: Tests skipped: Unable to retrieve first and last glyphs from cmap font.c:2549: Tests skipped: Font Droid Sans Japanese doesn't contain 'x', skipping the test font.c:2549: Tests skipped: Font Droid Sans Thai doesn't contain 'x', skipping the test font.c:2549: Tests skipped: Font OpenSymbol doesn't contain 'x', skipping the test font.c:2549: Tests skipped: Font rsfs10 doesn't contain 'x', skipping the test font.c:3756: Test failed: expected 19, got 0 (@IPAGothic:128) make[1]: *** [font.ok] Error 1 -- Alexandre Julliard julliard(a)winehq.org

2 1

Re: [PATCH] winebuild: Use Clang to assemble if found. (try 4)
by Alexandre Julliard Jan. 25, 2013

Jan. 25, 2013

Charles Davis <cdavis5x(a)gmail.com> writes: > @@ -346,13 +352,26 @@ struct strarray *get_as_command(void) > struct strarray *args = strarray_init(); > > if (!as_command) > + as_command = find_tool( "clang", NULL ); > + > + if (!as_command) > { > static const char * const commands[] = { "gas", "as", NULL }; > as_command = find_tool( "as", commands ); > } > + > + if (!as_command) > + fatal_error( "cannot find suitable assembler\n" ); > + > strarray_add_one( args, as_command ); > > - if (force_pointer_size) > + if (strstr( as_command, "clang" )) It's looking better, but strstr is not appropriate here. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: [PATCH 2/2] jscript: Fixed a leak in regexp_match function
by Marcus Meissner Jan. 25, 2013

Jan. 25, 2013

On Fri, Jan 25, 2013 at 04:26:59PM +0100, Piotr Caban wrote: > --- > dlls/jscript/regexp.c | 11 ++++++++--- > 1 file changed, 8 insertions(+), 3 deletions(-) > > > diff --git a/dlls/jscript/regexp.c b/dlls/jscript/regexp.c > index 0d66551..fa2668c 100644 > --- a/dlls/jscript/regexp.c > +++ b/dlls/jscript/regexp.c > @@ -3466,10 +3466,15 @@ static HRESULT regexp_match(script_ctx_t *ctx, jsdisp_t *dispex, jsstr_t *str, B > break; > > if(ret_size == i) { > - if(ret) > - ret = heap_realloc(ret, (ret_size <<= 1) * sizeof(match_result_t)); > - else > + if(ret) { > + match_result_t *old_ret = ret; > + > + ret = heap_realloc(old_ret, (ret_size <<= 1) * sizeof(match_result_t)); > + if(!ret) > + heap_free(old_ret); Should be probably if (ret) heap_free(old_ret); here. > + }else { > ret = heap_alloc((ret_size=4) * sizeof(match_result_t)); > + } > if(!ret) { > hres = E_OUTOFMEMORY; > break; > >

2 1