February 2015 - Wine-Devel - list.winehq.org

Re: msvcrt: Use Dragon4 algorithm to fix inconsistent rounding behaviour for sprintf and add tests. (try 2)

by Dmitry Timoshkov

YongHao Hu <christopherwuy(a)gmail.com> wrote: > +typedef int int32_t; > +typedef unsigned char uint8_t; > +typedef unsigned long uint32_t; > +typedef unsigned long long uint64_t; Please use appropriate win32 types instead. Besides using 'typedef unsigned long uint32_t' is wrong. -- Dmitry.

Feb. 27, 2015

Re: msvcrt: Use Dragon4 algorithm to fix inconsistent rounding behaviour for sprintf and add tests.

by YongHao Hu

Hi, Zhenbo. I have fixed the patch and send a new one to wine-patch, thank you. : ) The new patch looks good for me now. I switch to a new branch and use git apply ~/0012-msvcrt-Use-Dragon4-algorithm-to-fix-inconsistent-round.txt to apply my patch and compile it successfully. You can have a try. BTW, I had the author's consent to distribute his implementation using the following license, which I had added to files. >// Derived from code released under the ZLIB license >// Copyright (c) 2014 Ryan Juckett >// http://www.ryanjuckett.com/ On 2015年02月27日 09:32, Zhenbo Li wrote: > Hi, YongHao, > Quite a huge patch lol > > I tried your patch. > > $ patch -p1 <0012-msvcrt-Use-Dragon4-algorithm-to-fix-inconsistent-round.txt > (Stripping trailing CRs from patch; use --binary to disable.) > patching file dlls/msvcrt/dragon4.h > (Stripping trailing CRs from patch; use --binary to disable.) > patching file dlls/msvcrt/printf.h > (Stripping trailing CRs from patch; use --binary to disable.) > patching file dlls/msvcrt/tests/printf.c > (Stripping trailing CRs from patch; use --binary to disable.) > patching file dlls/msvcrt/wine_printfloat.h > > Is this critical? > > Also, your patch failed to compile on my computer. > > dragon4.h:10:0: error: unterminated #ifndef > #ifndef __WINE_STDIO_H > > Why is it necessary here? > > And then, wine_printfloat.h:26:15: error: unknown type name 'uint32_t' > It seems that you're using C99 types? It is not allowed in wine code[1] > > Finally, you've created two files with ZLIB license. Is it better to move all > the ZLIB license code to one file? I'm not sure about that. Could someone > have a check to avoid legal problems? > > > [1]: wiki.winehq.org/Developers-Hints > > 2015-02-26 23:13 GMT+08:00 YongHao Hu <christopherwuy(a)gmail.com>: >> Fix bug https://bugs.winehq.org/show_bug.cgi?id=37913 . >> The Dragon4 algorithm implementation is based on >> http://www.ryanjuckett.com/programming/printing-floating-point-numbers/. >> >> --- >> dlls/msvcrt/dragon4.h | 670 >> ++++++++++++++++++++++++++++++++++++++++++ >> dlls/msvcrt/printf.h | 43 +-- >> dlls/msvcrt/tests/printf.c | 83 +++++- >> dlls/msvcrt/wine_printfloat.h | 248 ++++++++++++++++ >> 4 files changed, 1026 insertions(+), 18 deletions(-) >> create mode 100644 dlls/msvcrt/dragon4.h >> create mode 100644 dlls/msvcrt/wine_printfloat.h >> >> >> >> >> > >

Feb. 27, 2015

Re: msvcrt: Use Dragon4 algorithm to fix inconsistent rounding behaviour for sprintf and add tests.

by Zhenbo Li

Hi, YongHao, Quite a huge patch lol I tried your patch. $ patch -p1 <0012-msvcrt-Use-Dragon4-algorithm-to-fix-inconsistent-round.txt (Stripping trailing CRs from patch; use --binary to disable.) patching file dlls/msvcrt/dragon4.h (Stripping trailing CRs from patch; use --binary to disable.) patching file dlls/msvcrt/printf.h (Stripping trailing CRs from patch; use --binary to disable.) patching file dlls/msvcrt/tests/printf.c (Stripping trailing CRs from patch; use --binary to disable.) patching file dlls/msvcrt/wine_printfloat.h Is this critical? Also, your patch failed to compile on my computer. dragon4.h:10:0: error: unterminated #ifndef #ifndef __WINE_STDIO_H Why is it necessary here? And then, wine_printfloat.h:26:15: error: unknown type name 'uint32_t' It seems that you're using C99 types? It is not allowed in wine code[1] Finally, you've created two files with ZLIB license. Is it better to move all the ZLIB license code to one file? I'm not sure about that. Could someone have a check to avoid legal problems? [1]: wiki.winehq.org/Developers-Hints 2015-02-26 23:13 GMT+08:00 YongHao Hu <christopherwuy(a)gmail.com>: > Fix bug https://bugs.winehq.org/show_bug.cgi?id=37913 . > The Dragon4 algorithm implementation is based on > http://www.ryanjuckett.com/programming/printing-floating-point-numbers/. > > --- > dlls/msvcrt/dragon4.h | 670 > ++++++++++++++++++++++++++++++++++++++++++ > dlls/msvcrt/printf.h | 43 +-- > dlls/msvcrt/tests/printf.c | 83 +++++- > dlls/msvcrt/wine_printfloat.h | 248 ++++++++++++++++ > 4 files changed, 1026 insertions(+), 18 deletions(-) > create mode 100644 dlls/msvcrt/dragon4.h > create mode 100644 dlls/msvcrt/wine_printfloat.h > > > > > -- Have a nice day! Zhenbo Li

Feb. 27, 2015

Re: [PATCH] [shell32] Implement SHRunControlPanel using ShellExecute

by Nikolay Sivov

On 26.02.2015 17:01, Robert Naumann wrote: > --- > dlls/shell32/shellord.c | 6 ++++-- > 1 file changed, 4 insertions(+), 2 deletions(-) > > diff --git a/dlls/shell32/shellord.c b/dlls/shell32/shellord.c > index 914dc5a..9ea3d62 100644 > --- a/dlls/shell32/shellord.c > +++ b/dlls/shell32/shellord.c > @@ -1129,8 +1129,10 @@ HRESULT WINAPI SHWinHelp (DWORD v, DWORD w, DWORD x, DWORD z) > */ > BOOL WINAPI SHRunControlPanel (LPCWSTR commandLine, HWND parent) > { > - FIXME("(%s, %p): stub\n", debugstr_w(commandLine), parent); > - return FALSE; > + if((int)ShellExecute(parent,NULL,commandLine,NULL,NULL,SW_NORMAL)>32) > + return TRUE; > + else > + return FALSE; > } Why 32? Also please don't remove tracing line.

Feb. 26, 2015

Re: [PATCH 1/1] handling sysex midi messages in winecoreaudio

by Andrew Eikum

On Wed, Feb 25, 2015 at 11:04:45PM +0100, tinez wrote: > Ok, i hope i got it right this time. > Yes, this looks OK to me. Send it off to wine-patches for inclusion, and please use your real, full name in the patch. Thanks! Andrew

Feb. 26, 2015

Re: [PATCH try2] urlmon: Added IInternetBindInfo implementation to BindStatusCallback object.

by Jacek Caban

On 02/26/15 11:10, Marvin wrote: > Hi, > > While running your changed tests on Windows, I think I found new failures. > Being a bot and all I'm not very good at pattern recognition, so I might be > wrong, but could you please double-check? > Full results can be found at > https://testbot.winehq.org/JobDetails.pl?Key=11751 > > Your paranoid android. > > > === w7u (32 bit url) === > url.c:3632: Test failed: expected OnProgress_DOWNLOADINGDATA > > === wvistau64 (64 bit url) === > url.c:3632: Test failed: expected OnProgress_DOWNLOADINGDATA These don't seem related. Jacek

Feb. 26, 2015

Re: [PATCH try2] urlmon: Added IInternetBindInfo implementation to BindStatusCallback object.

by Marvin

Hi, While running your changed tests on Windows, I think I found new failures. Being a bot and all I'm not very good at pattern recognition, so I might be wrong, but could you please double-check? Full results can be found at https://testbot.winehq.org/JobDetails.pl?Key=11751 Your paranoid android. === w7u (32 bit url) === url.c:3632: Test failed: expected OnProgress_DOWNLOADINGDATA === wvistau64 (64 bit url) === url.c:3632: Test failed: expected OnProgress_DOWNLOADINGDATA

Feb. 26, 2015

Re: [PATCH 3/5] wined3d: Remove ARB_/EXT_TEXTURE_ENV_ADD.

by Stefan Dösinger

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Am 2015-02-26 um 10:25 schrieb Henri Verbeet: > Right, we actually depend on at least texture_env_combine in > practice and that adds the ADD operation by itself. In theory we > could perhaps use texture_env_add when texture_env_combine isn't > available, but it doesn't seem worth it at this point. Fwiw, the only GPU I know of that doesn't at least have texture_env_combine is my ancient ATI Mach64 GPU. I agree that it's not worth the effort. The Mach64 DRM module never made it into the kernel, afaik the Mach64 DDX driver is abandoned and the the mesa parts were removed from Mesa 8.0. -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQIcBAEBAgAGBQJU7uyfAAoJEN0/YqbEcdMwtlYP/1g4d/E4g5dMNI7KinPYbx+Z 9MWhhUcegvC3rwNMMkJPSVSzR1WYbfDQgPCxuhKnMPuzNoa2ypviUuZArR/nhEZv +086ym1icpn0hAI+goDrL/DATcYyilHifM0gV5T64+4ieWMeNGrz8B4Jh3u9Bx3c lSJaKMH/e6Jos2Jp5MG37U97hSpdFsLL83RcQwXln67kMoucJOiqLvYnCDkzaD62 eAYZq8SucsupBGXNY0B6r6Q8/Z/geZsoiy6JzbFbvQgNJpnu1fvR3r4Bh8tejqOs P0WykVLOluHlnTuRMgaJYmr/WNOZAgWm+7YqoruIAlDgUAn9mGX01I6VwitI6xr8 zjoB4n4rhkMNCJy9uz6DC69IiRvqvEuB3KbshBUl6jGbBqxRjvGpz2AHgVzD2yy8 7whj9dtjq0wugI5saM2jSVuspSoVXxdO5igNL4RJLCA/IkPy8xwNvb1+dhzpchYv LAFPLu93Cy/OLA28mTcSfkwnwvbWhgSbBxyFhAaHfO8BDEjTJpDuULrnzs4q6KXb yvhHeIzXbBMVlMOk8PuIjOEEIyJhas7Nea4LJcOFyFs0CLNNpFNKq0UKsOm48Orj 0CHThLenyVwdlm3WDzXfPf2l84gLE8t6ev6SVKn+5EapfMEnDWKvHMkeL43sQ04j vfoFB4Ox+ZX6uZ35MAip =r+WW -----END PGP SIGNATURE-----

Feb. 26, 2015

Re: [PATCH 3/5] wined3d: Remove ARB_/EXT_TEXTURE_ENV_ADD.

by Henri Verbeet

On 25 February 2015 at 15:22, Matteo Bruni <matteo.mystral(a)gmail.com> wrote: > It's not all that clear and nice. set_tex_op() really depends on > GL_(ARB|EXT)_texture_env_combine, since it's always setting > GL_TEXTURE_ENV_MODE to GL_COMBINE (or GL_COMBINE4_NV) and using its > knobs to set the texture color or alpha operation state (and those > include the ADD setting for the COMBINE pname). Right, we actually depend on at least texture_env_combine in practice and that adds the ADD operation by itself. In theory we could perhaps use texture_env_add when texture_env_combine isn't available, but it doesn't seem worth it at this point.

Feb. 26, 2015

Re: [3/3] dpnet: Convert Address components to use a list

by Stefan Dösinger

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi Alistair, Alexandre made me aware that I missed a blaring issue here: Am 2015-02-26 um 04:32 schrieb Alistair Leslie-Hughes: > + temp = heap_realloc(This->components, sizeof(struct > component *) * This->comp_count+1 ); X * This->comp_count+1 won't do what you want it to do. Also he wants the array to be grown by doubling it. If the expected component count is low as you say, allocate the expected count at object creation (e.g. 4), and then double the array in case you need to grow it. Also, in terms of style, you can write sizeof(*This->components) instead of sizeof(struct component *). That way you won't have to change the code if you change the type of This->components. -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQIcBAEBAgAGBQJU7tiXAAoJEN0/YqbEcdMw6NwP/iWzcMCi48YkZ5iu52GrQDTo 86BL0M0VjFy72VDJq58pXvrK+7skm7JlI+jj/hdF0dtXg7ga6OCHqMwdyKdpUisB P93rCUVNOWvtqWltjct3UjdFM0PUD+IU9s5r3nX6sYMxwU++Kt31q5Cxq5APwxUO XplBpKeoRKtx44xgX0o84/hv6MijePDbocn+sQ3gBSyG4UdC4lt9F1Ze57r87DoW F0LkapFizu+45sRVWm6d6ANL4VL+pq3vVmjlIFRUPerVlIRRGEXnE3/5RqBegOUw MWFqRzj7LicQsy2tyan01o+OoUqGQL6ynesKj+I2LMf76i5JlSPGxoYMYlsk3vE+ JXn74dqP9iX6+Rhk7mv6q4iOQNOYsH5BraRWWLCT3vzr5DZESk+pjqn3ECKPkkZQ S88g9ismdDLkQ5pSNT/tbxeV2oRQSa4cTPp8aSNa5ttZtvwJhoHZjUGtdxAU7Go2 DcJuSbmTXFQo50Iaaw8Qlb+ZPbUVE0rjcALkKH74Bhq5vSbRkA1eu69A2TQOB9S3 x1sVVT78YhhWkl6qjESrhP0NFApNiCunUIK2U110aagOHGUyFaRix5DNN3lCO0ZT n9bvdXbZGCgg5RhxgIaqvTuGBhVZh882tpWecazipmYePUmO+4GzGwWGfxfyPymO qzUzMMRIGsukENMDUaus =3JhD -----END PGP SIGNATURE-----

Feb. 26, 2015