September 2017 - Wine-Devel

Re: [PATCH] wine/test.h: Increment correct counter for tests in todo blocks
by Nikolay Sivov Sept. 21, 2017

Sept. 21, 2017

On 21.09.2017 19:11, Fabian Maurer wrote: > Tests succeeding in todo blocks are counted as failures, > and failures as success. This patch fixes that. > > Signed-off-by: Fabian Maurer <dark.shadow4(a)web.de> > --- > include/wine/test.h | 4 ++-- > 1 file changed, 2 insertions(+), 2 deletions(-) > > diff --git a/include/wine/test.h b/include/wine/test.h > index 4ec66d5391..f0ee617d5f 100644 > --- a/include/wine/test.h > +++ b/include/wine/test.h > @@ -311,7 +311,7 @@ int winetest_vok( int condition, const char *msg, __winetest_va_list args ) > printf( "%s:%d: Test succeeded inside todo block: ", > data->current_file, data->current_line ); > vprintf(msg, args); > - InterlockedIncrement(&todo_failures); > + InterlockedIncrement(&todo_successes); > return 0; > } > else > @@ -322,7 +322,7 @@ int winetest_vok( int condition, const char *msg, __winetest_va_list args ) > data->current_file, data->current_line ); > vprintf(msg, args); > } > - InterlockedIncrement(&todo_successes); > + InterlockedIncrement(&todo_failures); > return 1; > } > } > This is correct. If test passes but is marked as todo it is a failure.

2 2

Re: [v5 1/3] evr: Add tests
by Alexandre Julliard Sept. 21, 2017

Sept. 21, 2017

Fabian Maurer <dark.shadow4(a)web.de> writes: > +static void test_evr_filter_aggregations(void) > +{ > + const IID * iids[] = { > + &IID_IAMCertifiedOutputProtection, &IID_IAMFilterMiscFlags, &IID_IBaseFilter, > + &IID_IKsPropertySet, &IID_IMediaEventSink, &IID_IMediaSeeking, &IID_IQualityControl, > + &IID_IQualProp, &IID_IEVRFilterConfig, &IID_IMFGetService, &IID_IMFVideoPositionMapper, > + &IID_IMFVideoRenderer, &IID_IQualityControl > + }; > + int i; > + > + if(!strcmp(winetest_platform, "wine")) > + { > + skip("Not supported yet.\n"); > + return; > + } I don't think it's useful to have a test that doesn't test anything. You should mark the tests as todo but still run them. If that's not possible, it should wait until we have something to test. -- Alexandre Julliard julliard(a)winehq.org

2 2

Re: [PATCH 3/3] msi: Defer the appropriate standard actions to the install script.
by Hans Leidekker Sept. 21, 2017

Sept. 21, 2017

On Wed, 2017-09-20 at 23:53 -0500, Zebediah Figura wrote: > diff --git a/dlls/msi/action.c b/dlls/msi/action.c > index 915ff173e9..b5ed7c75a8 100644 > --- a/dlls/msi/action.c > +++ b/dlls/msi/action.c > @@ -7764,6 +7764,7 @@ static UINT ACTION_InstallSFPCatalogFile( MSIPACKAGE *package ) > > static const struct > { > + BOOL deferred; > const WCHAR *action; > const UINT description; > const UINT template; > @@ -7772,83 +7773,83 @@ static const struct > } > StandardActions[] = > { [...] > + { 0, szAllocateRegistrySpace, IDS_DESC_ALLOCATEREGISTRYSPACE, IDS_TEMP_ALLOCATEREGISTRYSPACE, ACTION_AllocateRegistrySpace, NULL }, > + { 0, szAppSearch, IDS_DESC_APPSEARCH, IDS_TEMP_APPSEARCH, ACTION_AppSearch, NULL }, > + { 0, szBindImage, IDS_DESC_BINDIMAGE, IDS_TEMP_BINDIMAGE, ACTION_BindImage, NULL }, > + { 0, szCCPSearch, IDS_DESC_CCPSEARCH, 0, ACTION_CCPSearch, NULL }, > + { 0, szCostFinalize, IDS_DESC_COSTFINALIZE, 0, ACTION_CostFinalize, NULL }, > + { 0, szCostInitialize, IDS_DESC_COSTINITIALIZE, 0, ACTION_CostInitialize, NULL }, > + { 1, szCreateFolders, IDS_DESC_CREATEFOLDERS, IDS_TEMP_CREATEFOLDERS, ACTION_CreateFolders, szRemoveFolders }, > + { 1, szCreateShortcuts, IDS_DESC_CREATESHORTCUTS, IDS_TEMP_CREATESHORTCUTS, ACTION_CreateShortcuts, szRemoveShortcuts }, > + { 1, szDeleteServices, IDS_DESC_DELETESERVICES, IDS_TEMP_DELETESERVICES, ACTION_DeleteServices, szInstallServices }, > + { 0, szDisableRollback, 0, 0, ACTION_DisableRollback, NULL }, > + { 1, szDuplicateFiles, IDS_DESC_DUPLICATEFILES, IDS_TEMP_DUPLICATEFILES, ACTION_DuplicateFiles, szRemoveDuplicateFiles }, > + { 0, szExecuteAction, 0, 0, ACTION_ExecuteAction, NULL }, > + { 0, szFileCost, IDS_DESC_FILECOST, 0, ACTION_FileCost, NULL }, > + { 0, szFindRelatedProducts, IDS_DESC_FINDRELATEDPRODUCTS, IDS_TEMP_FINDRELATEDPRODUCTS, ACTION_FindRelatedProducts, NULL }, > + { 0, szForceReboot, 0, 0, ACTION_ForceReboot, NULL }, > + { 0, szInstallAdminPackage, IDS_DESC_INSTALLADMINPACKAGE, IDS_TEMP_INSTALLADMINPACKAGE, ACTION_InstallAdminPackage, NULL }, > + { 0, szInstallExecute, 0, 0, ACTION_InstallExecute, NULL }, > + { 0, szInstallExecuteAgain, 0, 0, ACTION_InstallExecute, NULL }, > + { 1, szInstallFiles, IDS_DESC_INSTALLFILES, IDS_TEMP_INSTALLFILES, ACTION_InstallFiles, szRemoveFiles }, > + { 0, szInstallFinalize, 0, 0, ACTION_InstallFinalize, NULL }, > + { 0, szInstallInitialize, 0, 0, ACTION_InstallInitialize, NULL }, > + { 1, szInstallODBC, IDS_DESC_INSTALLODBC, 0, ACTION_InstallODBC, szRemoveODBC }, > + { 1, szInstallServices, IDS_DESC_INSTALLSERVICES, IDS_TEMP_INSTALLSERVICES, ACTION_InstallServices, szDeleteServices }, > + { 0, szInstallSFPCatalogFile, IDS_DESC_INSTALLSFPCATALOGFILE, IDS_TEMP_INSTALLSFPCATALOGFILE, ACTION_InstallSFPCatalogFile, NULL }, > + { 0, szInstallValidate, IDS_DESC_INSTALLVALIDATE, 0, ACTION_InstallValidate, NULL }, > + { 0, szIsolateComponents, 0, 0, ACTION_IsolateComponents, NULL }, > + { 0, szLaunchConditions, IDS_DESC_LAUNCHCONDITIONS, 0, ACTION_LaunchConditions, NULL }, > + { 0, szMigrateFeatureStates, IDS_DESC_MIGRATEFEATURESTATES, IDS_TEMP_MIGRATEFEATURESTATES, ACTION_MigrateFeatureStates, NULL }, > + { 0, szMoveFiles, IDS_DESC_MOVEFILES, IDS_TEMP_MOVEFILES, ACTION_MoveFiles, NULL }, > + { 1, szMsiPublishAssemblies, IDS_DESC_MSIPUBLISHASSEMBLIES, IDS_TEMP_MSIPUBLISHASSEMBLIES, ACTION_MsiPublishAssemblies, szMsiUnpublishAssemblies }, > + { 1, szMsiUnpublishAssemblies, IDS_DESC_MSIUNPUBLISHASSEMBLIES, IDS_TEMP_MSIUNPUBLISHASSEMBLIES, ACTION_MsiUnpublishAssemblies, szMsiPublishAssemblies }, > + { 0, szPatchFiles, IDS_DESC_PATCHFILES, IDS_TEMP_PATCHFILES, ACTION_PatchFiles, NULL }, If the criterion is making modifications to the system, why are MoveFiles and PatchFiles not deferred?

2 1

Re: [PATCH 4/5] wined3d: Add texture flag for mipmap generation.
by Henri Verbeet Sept. 21, 2017

Sept. 21, 2017

On 20 September 2017 at 19:53, Matteo Bruni <mbruni(a)codeweavers.com> wrote: > Signed-off-by: Matteo Bruni <mbruni(a)codeweavers.com> > --- > Tests (later on in the patch series) show that calling GenerateMips() > on a texture without the D3D11_RESOURCE_MISC_GENERATE_MIPS flag does > nothing. In theory, the whole check could be done on the d3d11 side > but that turns out to be somewhat awkward, in practice, so I'm going > to add a wined3d flag and pass it around instead. > In principle that's fine, but like this it's unused. (For what it's worth, I don't necessarily think it needs to be used a lot initially. Even if it's only used to print the FIXME or not in wined3d_texture_generate_mipmaps() that would be a start.) There's also the matter of how this relates to WINED3DUSAGE_AUTOGENMIPMAP. This flag is for manual mipmap generation, while WINED3DUSAGE_AUTOGENMIPMAP is for automatic mipmap generation, but it's not pretty that one is a texture flag while the other is a resource usage flag.

1 0

Re: [PATCH 2/3] msi/tests: Add a helper to delete an installed package.
by Marvin Sept. 21, 2017

Sept. 21, 2017

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=33206 Your paranoid android. === w7u (32 bit install) === The task timed out

2 1

Re: ntdll: Implement set_cpu_context for ARM64 (try 2)
by Martin Storsjö Sept. 21, 2017

Sept. 21, 2017

On Wed, 20 Sep 2017, André Hentschel wrote: > Signed-off-by: André Hentschel <nerv(a)dawncrow.de> > --- > dlls/ntdll/signal_arm64.c | 41 +++++++++++++++++++++++++++++++++++++---- > 1 file changed, 37 insertions(+), 4 deletions(-) Now that you're implementing this, I think it'd make sense to complete the arm64 CONTEXT struct (compared to the win10 sdk); our version is lacking the float registers (and debug registers) at the end, when compared with the version of the struct in the public SDK. And contrary to ARM, support for the float registers is on by default in all toolchains, so there's no combination where one would be building wine for arm64 where that wouldn't work. Our version also defines the X registers slightly differently, the public SDK uses an union that allows accessing them both via member names and via an array - although it should end up as the same binary representation. I haven't sent any patch for bringing this up to date yet since I wasn't touching this area yet. // Martin

1 0

Re: cmd: Added support for "&&" and "||" condition checking.
by Alexandre Julliard Sept. 20, 2017

Sept. 20, 2017

Alexander Coffin <alexcoffin1999(a)gmail.com> writes: > @@ -1478,7 +1557,9 @@ void WCMD_execute (const WCHAR *command, const WCHAR *redirects, > switch (i) { > > case WCMD_CALL: > + push_errorlevel_changed(); > WCMD_call (p); > + pop_errorlevel_changed(); This seems to be the only place that does this, so I don't see why you need to manage your own stack instead of saving/restoring things on the process stack. Also returning a success/failure value from the various functions that implement commands would probably be cleaner than adding a global 'changed' flag. -- Alexandre Julliard julliard(a)winehq.org

2 1

How to proceed with shader errors for d3d11 game under wine
by pecisk＠gmail.com Sept. 20, 2017

Sept. 20, 2017

Hi! I am trying to find better way to collect and report information about game I want to make run under wine "Elite: Dangerous". At the launch of the game game compiles both regular shaders it is using in game and also compute shaders it's using for scenery generation. During that initial compilation game hangs and I see lot of these and similar errors: fixme:d3d_shader:shader_sm4_read_instruction_modifier Unhandled modifier 0x800002c2. fixme:d3d_shader:shader_sm4_read_instruction_modifier Unhandled modifier 0x00199983. fixme:d3d_shader:shader_sm4_read_instruction_modifier Unhandled modifier 0x800002c2. fixme:d3d_shader:shader_sm4_read_instruction_modifier Unhandled modifier 0x00199983. fixme:d3d_shader:shader_sm4_read_instruction_modifier Unhandled modifier 0x800002c2. fixme:d3d_shader:shader_sm4_read_instruction_modifier Unhandled modifier 0x00199983. fixme:d3d_shader:shader_sm4_read_instruction_modifier Unhandled modifier 0x800002c2. fixme:d3d_shader:shader_sm4_read_instruction_modifier Unhandled modifier 0x00199983. fixme:d3d_shader:shader_sm4_read_instruction_modifier Unhandled modifier 0x800002c2. fixme:d3d_shader:shader_sm4_read_instruction_modifier Unhandled modifier 0x00199983. fixme:d3d_shader:shader_sm4_read_instruction_modifier Unhandled modifier 0x800002c2. fixme:d3d_shader:shader_sm4_read_instruction_modifier Unhandled modifier 0x00199983. fixme:d3d_shader:shader_sm4_read_instruction_modifier Unhandled modifier 0x800002c2. fixme:d3d_shader:shader_sm4_read_instruction_modifier Unhandled modifier 0x00199983. fixme:d3d_shader:shader_sm4_read_instruction_modifier Unhandled modifier 0x800002c2. fixme:d3d_shader:shader_sm4_read_instruction_modifier Unhandled modifier 0x00199983. fixme:d3d_shader:print_glsl_info_log Info log received from GLSL shader #162: fixme:d3d_shader:print_glsl_info_log Compute info fixme:d3d_shader:print_glsl_info_log ------------ fixme:d3d_shader:print_glsl_info_log 0(22) : warning C7050: "R0.zw" might be used before being initialized fixme:d3d_shader:print_glsl_info_log 0(23) : warning C7050: "R1.zw" might be used before being initialized So question is - is there's a way to debug these more in detail? Are any such info useful for people working on d3d11 support? Thank you for any hints, Respectfully, Peteris Krisjanis.

1 0

Re: [v4 2/2] kernel32/tests: pipe: Added test for pipe security attributes
by Jacek Caban Sept. 20, 2017

Sept. 20, 2017

Hi Jonathan, For a quick test I made modification in your tests, see the attachment. With that modification, tests still pass on Windows but not on Wine (with or without your patches). This shows that there is something wrong with default security descriptor. I don't know what's the exact problem you're trying to fix, but it's likely that fixing default security descriptor is what would fix your problem, I would suggest to concentrate on that first. Thanks, Jacek

1 0

Re: ntdll: Implement set_cpu_context for ARM64 (try 2)
by Alexandre Julliard Sept. 20, 2017

Sept. 20, 2017

André Hentschel <nerv(a)dawncrow.de> writes: > + "ldr x15, [x0, #0x100]\n\t" > + "mov sp, x15\n\t" > + "ldr x15, [x0, #0x108]\n\t" > + "ldr x0, [x0, #0x8]\n\t" > + "br x15\n\t" This doesn't look very different from the previous try, and it still has the same problem. -- Alexandre Julliard julliard(a)winehq.org

1 0