- Wine-devel - list.winehq.org

Cuda Applications on Wine - suggesting to merge wine-staging/patches/nvcuda-CUDA_Support patches into upstream wine.
by Alois Schlögl 16 Nov '22

16 Nov '22

I'd like suggesting to merge the 17 patches from wine-staging/patches/nvcuda-CUDA_Support patches into upstream wine for the following reasons: Currently Cuda applications for Windows do not run in native Wine. Such an application is Warpem (available for Windows only), and CUDA-Z for Windows. I assume there are other applications as well. This could be fixed by pulling in the 17 patches wine-staging/patches/nvcuda-CUDA_Support. Possible disadvantages: - build dependency to cuda - some distros might not want to have a build dependency on cuda.h That's the choice of the package manager, but this should be endorsed and recommended for the above reason. - run-time dependency on cuda wine does not have a runtime dependency on cuda (probably due to dynamic linking) e.g. cuda-z can be started under wine, and stops gracefully with a message the cuda is not available. - number of APIs to maintain is currently done anyway (by the wine-staging team) From a the perspective of some application, Cuda is part of the operating system should be provided by the platform (Wine in our case). Overall, it seems merging the nvcuda-CUDA_Support into upstream Wine are the right thing to do. Best regards, Alois

7 8

Wine 8.0 code freeze
by Alexandre Julliard 14 Nov '22

14 Nov '22

Folks, With the end of the year approaching, it will soon be time to start the code freeze towards Wine 8.0. My plan is to have another two development releases, and start code freeze with Wine 7.23, on December 9. So if there are things that you want in 8.0, you have another 4 weeks to get them upstreamed... -- Alexandre Julliard julliard(a)winehq.org

1 0

Wine staging 7.21 release
by Alistair Leslie-Hughes 13 Nov '22

13 Nov '22

Binary packages for various distributions will be available from: https://www.winehq.org/download Summary since last release * Rebased to current wine 7.21 (495 patches are applied to wine vanilla) Upstreamed (Either directly from staging or fixed with a similar patch). * mshtml: Add IHTMLLocation::hash property's getter implementation. * uxtheme: Protect CloseThemeData() from invalid input. * user32/tests: Add some tests to see when MessageBox gains WS_EX_TOPMOST style. * user32: MessageBox should be topmost when MB_SYSTEMMODAL style is set. Added: * None Updated: * eventfd_synchronization * ntdll-Syscall_Emulation * ntdll-ForceBottomUpAlloc Where can you help * Run Steam/Battle.net/GOG/UPlay/Epic * Test your favorite game. * Test your favorite applications. * Improve staging patches and get them accepted upstream. * Suggest patches to be included in staging. As always, if you find a bug, please report it via https://bugs.winehq.org Best Regards Alistair.

3 2

Upcoming vkd3d 1.6 release.
by Henri Verbeet 10 Nov '22

10 Nov '22

Hi all, We'd like to release vkd3d 1.6 in about 4 weeks. As we're nearing that date, the vkd3d tree is going to become progressively frozen, starting today. Now is the time to test your applications for any regressions, review any API changes since vkd3d 1.5, review and improve the API documentation, and so on. The next release, vkd3d 1.7, is tentatively planned for late March next year. Regards, Henri

1 0

Winelib stack overflow
by Vitor Ramos 07 Nov '22

07 Nov '22

Hello, I am not sure if this is the appropriate channel but here we go. I am trying to port a windows c++ application using winelib and I found an issue while using some c++ STL containers. Here is a minimal example: # winelib.dll.so ```c++ #include <windows.h> #include <unordered_map> #include <string> extern "C" int f1(long z) { std::unordered_map<int, std::string> y; return 10; } spec file ```c++ 2 stdcall f1(long) ``` # Building winelib.dll.so winemaker --console --dll . winebuild --implib -E winedll.spec -o libwinedll.a winebuild --fake-module --dll -E winedll.spec -o winedll.dll make cp libwinedll.a winedll.dll.so /usr/local/lib/wine/x86_64-unix cp winedll.dll /home/$USER/.wine/drive_c/windows/system32 win64 app ```c++ #include <iostream> extern "C"{ int f1(long z); } int main(){ int ret = f1(2); std::cout << ret << std::endl; } ``` # Building win64 app winemaker . -iwinedll make # Running wine64 win64app.exe.so # Output 0110:err:virtual:virtual_setup_exception stack overflow 2128 bytes in thread 0110 addr 0x17008c804 stack 0x207b0 (0x20000-0x21000-0x120000) If I remove the "std::unordered_map<int, std::string> y" from the custom lib it runs normally. I suspect that wine uses too much of the stack to construct the object, but it's just a wild guess. Does anybody know what the issue could be? Is it possible to increase the stack size? Thanks.

3 4

TestBot News
by Francois Gouget 07 Nov '22

07 Nov '22

* The TestBot has a new Windows 11 VM. It's a new Windows version so it has almost 40 extra failures. To avoid false positives it's only going to be running WineTest until the new failures have been flagged. But you can already select it in the list of available VMs to run tests on it. * There are now two PCI-Passthrough configurations: - w11pro64_amd -> AMD RX6600 GPU - w11pro64_nv -> NVIDIA RTX3050 GPU These have a surprising number of extra failures, about 12, when fgtb-w10pro64-rx550 only has about 2 extra failures. So I will recheck the configuration of the new VMs (on the AMD side it may have the clipboard problem again, but then a number of failures appeared to be in the d3d tests). These are also only running WineTest for now so as to not cause false positives but you can select them in the list of available VMs. -- Francois Gouget <fgouget(a)codeweavers.com>

1 2

invalid read errno bug
by xantares09＠hotmail.com 01 Nov '22

01 Nov '22

Hello, I'm hoping to get more attention for this bug report that have not been confirmed yet: https://bugs.winehq.org/show_bug.cgi?id=53677 Everything should be provided to ease reproducing, Regards

1 0

Wine staging 7.20 release
by Alistair Leslie-Hughes 01 Nov '22

01 Nov '22

Binary packages for various distributions will be available from: https://www.winehq.org/download Summary since last release * Rebased to current wine 7.20 (504 patches are applied to wine vanilla) Upstreamed (Either directly from staging or fixed with a similar patch). * sapi: Implement ISpObjectToken OpenKey * sapi: EnumTokens setup enumeration members * sapi: Implement ISpObjectTokenEnumBuilder Item * sapi: Add default voice registry key * sapi: Return dump object in ISpObjectTokenEnumBuilder Next Added: * None Updated: * nvcuda-CUDA_Support * mfplat-streaming-support Where can you help * Run Steam/Battle.net/GOG/UPlay/Epic * Test your favorite game. * Test your favorite applications. * Improve staging patches and get them accepted upstream. * Suggest patches to be included in staging. As always, if you find a bug, please report it via https://bugs.winehq.org Best Regards Alistair.

1 0

Re: ntdll: Fix build on systems without ENODATA (was: [PATCH] ntdll: Complete error case in get_initial_environment)
by Erich E. Hoover 19 Oct '22

19 Oct '22

On Wed, Oct 19, 2022 at 3:05 PM Gerald Pfeifer <gerald(a)pfeifer.com> wrote: > > On Wed, 19 Oct 2022, Erich E. Hoover wrote: > >> https://gitlab.winehq.org/wine/wine/-/merge_requests/1097 > > It looks like this just got merged. Gerald, would you check and see > > if ENOATTR is defined on your system? > > I checked, and FreeBSD (12 and 14, so current and future versions) both > define ENOATTR guarded by #ifndef _POSIX_SOURCE, so generally available. > > > This commit removes the compile warning > > It actually was hard build error. Yes, sorry - I misspoke. > > but you're going to get a lot of unnecessary console warnings > > ("attribute does not exist" is not an error here). > > Is it possible Alexandre's adjusted version of my submission avoids this? > ... No. If the filesystem supports extended attributes (!ENOTSUP) then you will get an error if there aren't any xattr for the file. I suspect that we either want: === #ifdef ENODATA if (errno == ENODATA) return ret; #endif #ifdef ENOATTR if (errno == ENOATTR) return ret; #endif === Or we want to do something like this: === #ifndef ENOATTR #define ENOATTR ENODATA #endif === and then check against ENOATTR instead (little bit of research says that platforms that define both use ENOATTR in this case). Best, Erich

1 0

Re: [PATCH] ntdll: Complete error case in get_initial_environment
by Esme Povirk (any/all) 19 Oct '22

19 Oct '22

> That's more involved than the previous process, but I think I got it: > https://gitlab.winehq.org/gerald/wine/-/commit/2bdbfb583452a62c7d6acf7f416b… > > Is there anything else to do? I found the actual MR at https://gitlab.winehq.org/wine/wine/-/merge_requests/1097, and it looks right to me. Nothing else to do for now, hopefully someone will come along and review it. > PS: It shows the test-linux32 pipeline as failed, while the other two > succeeded; looks like a false positive to the, though. Yes, those errors have been showing up for every MR. -- https://my.pronouns.page/are/any

2 1