April 2011 - Wine-devel - list.winehq.org

Re: winex11.drv: Ignore the zero-size owners used by Delphi apps when deciding whether to add NET_WM_STATE_SKIP_TASKBAR.
by Alexandre Julliard 21 Apr '11

21 Apr '11

Dmitry Timoshkov <dmitry(a)codeweavers.com> writes: > This patch reverts a part of 61e50e15ba45ad54655f98619f5ef33917033165 and > fixes the regression reported in the bug 26670. This will cause spurious task bar entries in many apps. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: [PATCH 1/4] msvcp90: Added basic_string<char>::append(size_t, char) implementation.
by Piotr Caban 21 Apr '11

21 Apr '11

Hi, On 04/18/11 23:53, Vincas Miliūnas wrote: > My remark: (The method is implemented by reusing existing code) this->size+count<this->size is the same as count< 0 and throwing a "string too long" exception in that case is not semantic. count is an unsigned variable so it cannot be smaller then 0. My intention was to make sure there's no overflow. This check is not needed because MSVCP_basic_string_char_npos-this->size <= count already makes sure this->size+count has small enough value. +basic_string_wchar* __thiscall MSVCP_basic_string_wchar_append_len_wchar( + basic_string_wchar *this, size_t count, wchar_t c) +{ + TRACE("%p %lu %lc\n", this, (unsigned long)count, c); It's probably not allowed to use %lc format. Cheers, Piotr

2 1

Re: d3dxof/tests: Make expected value in string and comparison match.
by Marvin 20 Apr '11

20 Apr '11

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 http://testbot.winehq.org/JobDetails.pl?Key=10525 Your paranoid android. === WVISTAADM (32 bit d3dxof) === Failure running script in VM: The specified guest user must be logged in interactively to perform this operation

1 0

Re: msvcrt: Added _resetstkoflw implementation
by Alexandre Julliard 20 Apr '11

20 Apr '11

Piotr Caban <piotr(a)codeweavers.com> writes: > +/********************************************************************* > + * _resetstkoflw (MSVCRT.@) > + */ > +int CDECL _resetstkoflw(void) > +{ > + MEMORY_BASIC_INFORMATION mbi; > + SYSTEM_INFO info; > + DWORD prot; > + void *alloc; > + > + GetSystemInfo(&info); > + if(!VirtualQuery(&mbi, &mbi, sizeof(mbi))) > + return 0; > + > + alloc = VirtualAlloc((char*)mbi.BaseAddress, info.dwPageSize, MEM_COMMIT, PAGE_READWRITE); > + if(!alloc) > + return 0; > + > + return VirtualProtect(alloc, info.dwPageSize, PAGE_GUARD|PAGE_READWRITE, &prot); This doesn't make sense to me, why do you need a VirtualAlloc in there? And what about the limits in the TEB? -- Alexandre Julliard julliard(a)winehq.org

2 2

Re: [PATCH 03/10] msvcrt: Use callback based printf in fprintf functions family (try2)
by Alexandre Julliard 19 Apr '11

19 Apr '11

Piotr Caban <piotr(a)codeweavers.com> writes: > --- > dlls/msvcrt/file.c | 118 > ++++++++++------------------------------------------ > 1 files changed, 22 insertions(+), 96 deletions(-) This one is causing console apps to output null characters. You'll have to spend more time on this patch series, it doesn't seem quite ready. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: include: Add rmxftmpl.h with contents generated from rmxftmpl.x.
by Alexandre Julliard 19 Apr '11

19 Apr '11

Dylan Smith <dylan.ah.smith(a)gmail.com> writes: > +#ifndef __WINE_RMXFTMPL_H > +#define __WINE_RMXFTMPL_H > + > +/* sizeof(D3DRM_XTEMPLATES) */ > +#define D3DRM_XTEMPLATE_BYTES 3278 > + > +/* Contents of D3DRM_XTEMPLATES generated using: > + * ../tools/binencxftmpl rmxftmpl.x | xxd -i > + */ You should generate the whole file, it shouldn't need to be updated by hand. -- Alexandre Julliard julliard(a)winehq.org

2 1

Re: serial: Fix race for IOCTL_SERIAL_WAIT_ON_MASK
by Alexandre Julliard 19 Apr '11

19 Apr '11

Wolfgang Walter <wolfgang.walter(a)stwm.de> writes: > set status field of piosb to STATUS_PENDING before calling wait_on(). If wait_on > returns with STATUS_PENDING don't touch piosb. > > Reason: if wait_on returns with STATUS_PENDING it started a thread which > itself modifies the status field. In general the iosb is not modified until the operation completes. -- Alexandre Julliard julliard(a)winehq.org

2 3

Cairo Graphics
by Jaime Rave 19 Apr '11

19 Apr '11

Hey Guys, On some of the bugs that i have reported to freedesktop, i have been asked about why Wine is making calls like XDrawSegmentand not using Cairo to handle the 2d drawing which always provides consistent behavior regardless of drivers[1]. I have also read that cairo support "X Window System, Quartz, Win32, image buffers, PostScript, PDF, and SVG file output. Experimental backends includeOpenGL <http://www.cairographics.org/OpenGL/>, XCB, BeOS, OS/2, and DirectFB"[2]. So it will also works with Wayland when this is ready :D. So i think this is a good way to make Wine more graphic driver agnostic and improve the behavior on MacOS without having to write a complete new driver just for Quartz. So i would like to know your thoughts, if Cairo could improve some stuff or if is not related or useless to the project Greetings Jaime Rave [1] https://bugs.freedesktop.org/show_bug.cgi?id=22770#c14 [2] http://www.cairographics.org/

2 1

Re: [PATCH 06/10] msvcrt: Added sprintf_p_l implementation (try2)
by Alexandre Julliard 19 Apr '11

19 Apr '11

Piotr Caban <piotr(a)codeweavers.com> writes: > @@ -552,6 +552,74 @@ int FUNC_NAME(pf_printf)(FUNC_NAME(puts_clbk) pf_puts, void *puts_ctx, const API > return written; > } > > +#ifndef PRINTF_WIDE > +/* This functions stores sizes of arguments. It uses args[0] > + * internally (in the following way): > + * 0x1 - using positional parameters > + * 0x2 - not using positional parameters > + * 0x4 - error occurred > + */ > +static printf_arg arg_clbk_size(void *ctx, int pos, int type, __ms_va_list *valist) > +{ > + static const printf_arg ret; > + printf_arg *args = ctx; > + > + if(pos == -1) { > + args[0].get_int |= 1; > + return ret; > + } else > + args[0].get_int |= 2; > + > + if(pos<1 || pos>MSVCRT__ARGMAX || (args[pos].get_int && args[pos].get_int!=type)) > + args[0].get_int |= 4; > + else > + args[pos].get_int = type; > + return ret; > +} Using magic values like this is ugly, please add proper defines. Also you probably want to fix the name of the function and the comments. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: [PATCH] Fix SystemProcessorPerformanceInformation to include DPC and Interrupt times
by Alexandre Julliard 19 Apr '11

19 Apr '11

"Ray Hinchliffe (RH)" <ray(a)rh-software.com> writes: > 1) When possible report DPC and Interrupt times > > 2) Fix _APPLE_ to return times in 100ns units and zero the unset members > > 3) Change the failure code to return 100ns values > > 4) Expunge unnecessary use of RtlAllocateHeap() etc.. Separate changes should be separate patches. > @@ -1241,8 +1241,9 @@ typedef struct _SYSTEM_PROCESSOR_PERFORMANCE_INFORMATION { > LARGE_INTEGER IdleTime; > LARGE_INTEGER KernelTime; > LARGE_INTEGER UserTime; > - LARGE_INTEGER Reserved1[2]; > - ULONG Reserved2; > + LARGE_INTEGER DpcTime; > + LARGE_INTEGER IntTime; > + ULONG IntCount; You can't change the public definition. -- Alexandre Julliard julliard(a)winehq.org

1 0