April 2017 - Wine-devel - list.winehq.org

Re: [PATCH] winhttp/tests: Test when large amount of data is returned from a 410
by Marvin 11 Apr '17

11 Apr '17

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=30136 Your paranoid android. === wvistau64 (32 bit winhttp) === winhttp.c:1143: Test failed: request failed unexpectedly 400 === w2008s64 (64 bit winhttp) === winhttp.c:1143: Test failed: request failed unexpectedly 400

1 0

Re: [PATCH 2/5] wined3d: Send texture dirty region updates through the command stream.
by Józef Kucia 11 Apr '17

11 Apr '17

On Tue, Apr 11, 2017 at 11:24 PM, Henri Verbeet <hverbeet(a)codeweavers.com> wrote: > - context = context_acquire(texture->resource.device, NULL, 0); > - sub_resource_idx = layer * texture->level_count; > - for (i = 0; i < texture->level_count; ++i, ++sub_resource_idx) > - { > - if (!wined3d_texture_load_location(texture, sub_resource_idx, context, texture->resource.map_binding)) > - { > - ERR("Failed to load location %s.\n", wined3d_debug_location(texture->resource.map_binding)); > - context_release(context); > - return E_OUTOFMEMORY; > - } > - wined3d_texture_invalidate_location(texture, sub_resource_idx, ~texture->resource.map_binding); > - } > - context_release(context); > + wined3d_cs_emit_add_dirty_texture_region(texture->resource.device->cs, texture, layer * texture->level_count); I think that just "layer" should be passed as the last argument to wined3d_cs_emit_add_dirty_texture_region().

1 0

Re: [PATCH v2] mountmgr: Create registry entries for serial ports.
by Alexandre Julliard 11 Apr '17

11 Apr '17

Alex Henrie <alexhenrie24(a)gmail.com> writes: > @@ -453,7 +457,21 @@ NTSTATUS WINAPI DriverEntry( DRIVER_OBJECT *driver, UNICODE_STRING *path ) > RegCloseKey( hkey ); > if (!RegCreateKeyExW( devicemap_key, serialW, 0, NULL, REG_OPTION_VOLATILE, > KEY_ALL_ACCESS, NULL, &hkey, NULL )) > + { > + /* create a registry entry for each serial port */ > + for (i = 1; i <= 256; i++) > + { > + sprintfW( pathW, comW, i ); > + serial_device = CreateFileW( pathW, 0, 0, NULL, OPEN_EXISTING, FILE_ATTRIBUTE_NORMAL, NULL ); > + if (serial_device == INVALID_HANDLE_VALUE) > + break; > + > + CloseHandle( serial_device ); > + RegSetValueExW( hkey, pathW + 4, 0, REG_SZ, (BYTE *)(pathW + 4), > + (strlenW( pathW + 4 ) + 1) * sizeof(WCHAR) ); The registry key name is supposed to be the NT device. Ideally, what you'd want is to actually create the NT device pointing it to the corresponding Unix device, as well as the DOS symlink. If you do this, CreateFile can then use the normal path, and the ntdll serial device detection logic can be moved to mountmgr. -- Alexandre Julliard julliard(a)winehq.org

1 0

Putting in place traffic shaping on winehq
by Jeremy White 11 Apr '17

11 Apr '17

Hey folks, We've been getting nailed with peak charges on the traffic coming from the WineHQ web server. We think a large reason for this is that we were not sophisticated in how we flushed the fastly cache, so we'd get slammed in a big rush when we flushed. I believe that Jeremy Newman has fixed that now, so hopefully it won't happen again. But those overcharges are brutal (i.e. thousands of dollars) :-(. So I'm planning to put in place some traffic shaping to provide a hard upper limit for winehq (I'm going to impose 8Mbit; we currently pay for 10Mbpit committed; if I up that to 20, I may expand that). This should not disrupt anything, but I thought I'd warn folks, so you'd know to blame me if things suddenly go dark <grin>. Cheers, Jeremy

3 4

Re: [PATCH 2/4] user32/tests: Correctly restore the caption sizes when run in DPI scaling mode.
by Marvin 11 Apr '17

11 Apr '17

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=30112 Your paranoid android. === w8 (32 bit sysparams) === sysparams.c:2500: Test failed: Waiting for the WM_DISPLAYCHANGE message timed out sysparams.c:2511: Test failed: Set bpp 32, but WM_DISPLAYCHANGE reported bpp -1 === w864 (32 bit sysparams) === sysparams.c:272: SPI_SETSCREENSAVETIMEOUT failed for reason: 329. Indicating test failure and skipping remainder of test sysparams.c:273: Test failed: SystemParametersInfoA: rc=0 err=329 sysparams.c:273: Test failed: SystemParametersInfoA: rc=0 err=329 sysparams.c:2500: Test failed: Waiting for the WM_DISPLAYCHANGE message timed out sysparams.c:2511: Test failed: Set bpp 32, but WM_DISPLAYCHANGE reported bpp -1

2 1

Re: [PATCH 4/4] user32/tests: Fix tests for minimium icon spacing when not in DPI scaling mode.
by Marvin 11 Apr '17

11 Apr '17

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=30117 Your paranoid android. === w8 (32 bit sysparams) === sysparams.c:2500: Test failed: Waiting for the WM_DISPLAYCHANGE message timed out sysparams.c:2511: Test failed: Set bpp 32, but WM_DISPLAYCHANGE reported bpp -1

1 0

Re: [PATCH 3/4] user32/tests: Fix a few failures in DPI scaling mode.
by Marvin 11 Apr '17

11 Apr '17

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=30116 Your paranoid android. === w7u (32 bit sysparams) === sysparams.c:246: Test failed: Missed a message: change_counter=0 sysparams.c:249: Test failed: Wrong action got 0 expected 4 sysparams.c:246: Test failed: Missed a message: change_counter=0 sysparams.c:249: Test failed: Wrong action got 0 expected 4 === w8 (32 bit sysparams) === sysparams.c:2500: Test failed: Waiting for the WM_DISPLAYCHANGE message timed out sysparams.c:2511: Test failed: Set bpp 32, but WM_DISPLAYCHANGE reported bpp -1 === w864 (32 bit sysparams) === The task timed out

1 0

Re: [PATCH 2/2] regedit: Re-implement convertHextToDWord() (v3)
by Alexandre Julliard 11 Apr '17

11 Apr '17

Hugh McMaster <hugh.mcmaster(a)outlook.com> writes: > + while (*buf == ' ' || *buf == '\t') buf++; > + if (!*buf) goto exit; > + > + p = buf; > + while (isxdigit(*p)) > + { > + count++; > + p++; > } > - return TRUE; > + if (count > 8) goto exit; > + > + end = p; > + while (*p == ' ' || *p == '\t') p++; > + if (*p && *p != ';') goto exit; > + > + *end = 0; > + *dw = strtoul(buf, &end, 16); > + ret = TRUE; > + > + exit: > + if (!ret) output_message(STRING_INVALID_HEX); > + HeapFree(GetProcessHeap(), 0, buf); The pointer has been modified, you need to free the original one. -- Alexandre Julliard julliard(a)winehq.org

2 2

Re: [PATCH 1/2] endproc.c: Implement ProcessPage_OnEndProcessTree
by Stefan Dösinger 11 Apr '17

11 Apr '17

Hi Akarsha, Thank you for your patches! My apologies for the very late review of your code. I am currently catching up on my mail after moving :-\ . Your patch has numerous compile errors. Please make sure that the code compiles without errors or warnings before sending it. Am 2017-03-30 um 19:57 schrieb Akarsha Sehwag: > +typedef struct process_list { > + int *pid; /*dynamic array to store the process IDs*/ MODULEENTRY32.th32ProcessID is a DWORD, please use it here as well. > + SIZE_T count; /*index to maintain the last entry of the array;*/ > + SIZE_T size; /*the current size of the pid array*/ > +} process_list; Is it necessary to store the child processes in a separate list? I'd expect that CreateToolhelp32Snapshot creates a snapshot that is not affected by changes to the process tree after it returns. If this is the case you can kill processes as you iterate over the snapshot with Process32Next. This would simplify the code a lot If you keep the list please use a wine list from include/wine/list.h. This will also give you more flexibility in the order of how you kill the processes, see below. > +static void init_process_list(process_list *list) { > + list->size = 4; /*initialise size with 4. Will increase if necessary.*/ > + list->pid = HeapAlloc(GetProcessHeap(), HEAP_ZERO_MEMORY, size * sizeof(int)); Minor style hint: You can use sizeof(*list->pid) instead of sizeof(int). That way the code doesn't need changes if the type of pid changes. > + if(!Process32First(snapshot, &entry)) > + return; > + > + do > + { > + if(entry.th32ParentProcessID == pid) > + process_list_append(list, entry.th32ProcessID); > + } while (Process32Next(snapshot, &entry)); > + > + end = list->count; > + > + for(i = start; i < end; ++i) > + { > + enum_process_children(snapshot, list, list->pid[i]); > + } I am not sure if this matters, but this will kill parent processes before their children. It may or may not be the correct thing and may be an interesting thing to find out with a test (unfortunately writing automated tests for taskmgr is not easy) I think the same kind of recursion should work if you operate on the snapshot directly without the helper list. > + for(i = 0; i <= list->last; ++i) { > + hProcess = OpenProcess(PROCESS_TERMINATE, FALSE, list->pid[i]); > > --- patch 2 --- > - for(i = 0; i <= list->last; ++i) { > + for(i = 0; i < list->count; ++i) { This should be merged into patch 1. It is important that the code always builds to prevent problems in git bisect (e.g. assume I am looking for a regression in the d3d code, but one of the revisions I want to test doesn't compile taskmgr) With best regards, Stefan

1 1

Help on how to run winetest in debug
by Yashodeep Mahapatra 11 Apr '17

11 Apr '17

Hi All, This is Yashodeep Mahapatra from Birla Institute Of Technology ,Mesra,India. I tried running wine on my machine after building and installing it. But the problem that I am facing is that my system just keeps running the process wine and displays Wine Shell Script in the window. Running wine winecfg also runs in the same way without giving any output on the screen. I also tried running notepad.exe but only a blank screen appeared. wine winecfg command led to this output: err:menubuilder:convert_to_native_icon error 0x80004005 getting frame 0 err:wincodecs:PngDecoder_CreateInstance Failed reading PNG because unable to find libpng16.16.dylib err:menubuilder:convert_to_native_icon error 0x80004005 getting frame 0 err:wincodecs:PngDecoder_CreateInstance Failed reading PNG because unable to find libpng16.16.dylib err:menubuilder:convert_to_native_icon error 0x80004005 getting frame 0 err:wincodecs:PngDecoder_CreateInstance Failed reading PNG because unable to find libpng16.16.dylib err:menubuilder:convert_to_native_icon error 0x80004005 getting frame 0 and nothing appeared on the screen. wine /programs/winetest/winetest.exe.so gave the following result: err:wincodecs:PngDecoder_CreateInstance Failed reading PNG because unable to find libpng16.16.dylib err:menubuilder:convert_to_native_icon error 0x80004005 getting frame 0 err:wincodecs:PngDecoder_CreateInstance Failed reading PNG because unable to find libpng16.16.dylib err:menubuilder:convert_to_native_icon error 0x80004005 getting frame 0 Can anyone please help me?

3 2