This goes on top of MR 320 and 345.
--
v6: tests/shader-runner: Add a '--dump-dxil' command line switch.
tests/shader-runner: Test shaders with dxcompiler.
tests/shader-runner: Replace immediate shader type strings with an enum.
tests/shader-runner: Do not exit if a 'require' directive is not met.
tests/shader-runner: Handle individual keywords in shader directives.
vkd3d-shader/dxil: Implement DX instruction LoadInput.
vkd3d-shader/dxil: Declare shader …
[View More]inputs.
vkd3d-shader/dxbc: Load input signatures also from ISG1 chunks.
vkd3d-shader/spirv: Build undefined values once.
vkd3d-shader/spirv: Introduce a Static Single Assignment register type.
vkd3d-shader/d3d-asm: Trace undefined registers.
vkd3d-shader: Make the paramater allocator slab size at least MAX_REG_OUTPUT.
https://gitlab.winehq.org/wine/vkd3d/-/merge_requests/346
[View Less]
These are the tests from !3367 (Defer listens on addresses currently bound first by another socket). The tests marked `todo_wine` rely on a more accurate implementation of `SO_REUSEADDR`.
--
https://gitlab.winehq.org/wine/wine/-/merge_requests/3921
Windows 10 [received support](https://devblogs.microsoft.com/commandline/af_unix-comes-to-window… for AF_UNIX sockets in Insider Build 17063. This merge request adds basic support for AF_UNIX sockets to ws2_32 and wineserver.
Of particular note is the difficulty in handling `sun_path`. Most of the functions that allow for translating Windows paths to Unix paths are not accessible from ws2_32. I considered the following options:
* Pass the Windows path to wineserver and do the conversion there.…
[View More]
* This is, as far as I can tell, not possible without major rearchitecting. wineserver does not have functions to translate Windows paths to Unix paths, for obvious reasons.
* Obtain the current working directory of the requesting process and temporarily change directories to there.
* This only handles relative paths and fails for absolute paths, UNC paths, etc.
* Conditionally change directories based on whether the path is relative or not.
* This is error-prone and wineserver does not have the requisite functions to do this cleanly.
I ultimately decided to pass the translated Unix path to wineserver, which changes directories to `dirname(path)`. It then provides `bind` and `connect` with `basename(path)`. This is not threadsafe, but wineserver is not (currently) multithreaded.
Abstract sockets are not fully supported by this patch, matching the behavior of Windows.
--
v59: ws2_32/tests: Add test for AF_UNIX sockets.
server: Fix getsockname() and accept() on AF_UNIX sockets.
server: Introduce error when attempting to create a SOCK_DGRAM AF_UNIX socket.
server: Allow for deletion of socket files.
ws2_32: Add support for AF_UNIX sockets.
ws2_32: Add afunix.h header.
https://gitlab.winehq.org/wine/wine/-/merge_requests/2786
[View Less]
Signed-off-by: Nikolay Sivov <nsivov(a)codeweavers.com>
--
v2: d3d10/effect: Add signed integer compare instructions for expressions.
d3d10/effect: Implement instructions used for signed integer division.
https://gitlab.winehq.org/wine/wine/-/merge_requests/3896
On Thu Sep 21 20:34:12 2023 +0000, Aidan Khoury wrote:
> I can for this, but we also need to check if it's a 32-bit image since
> patch transforms are never applied to 64-bit images
`RtlImageNtHeader` lacks proper image boundary checks for e_lfanew, so using it in this case is not safe.
--
https://gitlab.winehq.org/wine/wine/-/merge_requests/3870#note_46247
This clears up much of the todo list for mspatcha. GetFilePatchSignature* and NormalizeFileForPatchSignature have now been implemented in these changes.
These changes bring better support for things like the Adobe Acrobat installer without the need for things like the winetricks mspatcha.dll native dll override. Still needed is support for interleaved streams in the LZXD decompression logic, however this is a great start to better supporting software installers that use the Windows interface …
[View More]for creating and applying patches to files.
This is the beginning of the fixes required for bug 12501: https://bugs.winehq.org/show_bug.cgi?id=12501
--
v32: mspatcha: Use unaligned typedefs for basic types
mspatcha: Use string comparison for section names
mspatcha: Add support for 32-bit file image patch transforms
mspatcha: Add implementations for GetFilePatchSignature* routines
mspatcha: Add support for 32-bit file normalization
https://gitlab.winehq.org/wine/wine/-/merge_requests/3870
[View Less]
This clears up much of the todo list for mspatcha. GetFilePatchSignature* and NormalizeFileForPatchSignature have now been implemented in these changes.
These changes bring better support for things like the Adobe Acrobat installer without the need for things like the winetricks mspatcha.dll native dll override. Still needed is support for interleaved streams in the LZXD decompression logic, however this is a great start to better supporting software installers that use the Windows interface …
[View More]for creating and applying patches to files.
This is the beginning of the fixes required for bug 12501: https://bugs.winehq.org/show_bug.cgi?id=12501
--
v31: mspatcha: Use unaligned typedefs for basic types
mspatcha: Use string comparison for section names
mspatcha: Relocate PE/COFF image functions
mspatcha: Add support for 32-bit file image patch transforms
mspatcha: Add implementations for GetFilePatchSignature* routines
mspatcha: Add support for 32-bit file normalization
https://gitlab.winehq.org/wine/wine/-/merge_requests/3870
[View Less]
This clears up much of the todo list for mspatcha. GetFilePatchSignature* and NormalizeFileForPatchSignature have now been implemented in these changes.
These changes bring better support for things like the Adobe Acrobat installer without the need for things like the winetricks mspatcha.dll native dll override. Still needed is support for interleaved streams in the LZXD decompression logic, however this is a great start to better supporting software installers that use the Windows interface …
[View More]for creating and applying patches to files.
This is the beginning of the fixes required for bug 12501: https://bugs.winehq.org/show_bug.cgi?id=12501
--
v30: mspatcha: Use unaligned typedefs for basic types
mspatcha: Use string comparison for section names
https://gitlab.winehq.org/wine/wine/-/merge_requests/3870
[View Less]
This clears up much of the todo list for mspatcha. GetFilePatchSignature* and NormalizeFileForPatchSignature have now been implemented in these changes.
These changes bring better support for things like the Adobe Acrobat installer without the need for things like the winetricks mspatcha.dll native dll override. Still needed is support for interleaved streams in the LZXD decompression logic, however this is a great start to better supporting software installers that use the Windows interface …
[View More]for creating and applying patches to files.
This is the beginning of the fixes required for bug 12501: https://bugs.winehq.org/show_bug.cgi?id=12501
--
v29: mspatcha: Use string comparison for section names
mspatcha: Use unaligned typedefs for basic types
https://gitlab.winehq.org/wine/wine/-/merge_requests/3870
[View Less]
Windows 10 [received support](https://devblogs.microsoft.com/commandline/af_unix-comes-to-window… for AF_UNIX sockets in Insider Build 17063. This merge request adds basic support for AF_UNIX sockets to ws2_32 and wineserver.
Of particular note is the difficulty in handling `sun_path`. Most of the functions that allow for translating Windows paths to Unix paths are not accessible from ws2_32. I considered the following options:
* Pass the Windows path to wineserver and do the conversion there.…
[View More]
* This is, as far as I can tell, not possible without major rearchitecting. wineserver does not have functions to translate Windows paths to Unix paths, for obvious reasons.
* Obtain the current working directory of the requesting process and temporarily change directories to there.
* This only handles relative paths and fails for absolute paths, UNC paths, etc.
* Conditionally change directories based on whether the path is relative or not.
* This is error-prone and wineserver does not have the requisite functions to do this cleanly.
I ultimately decided to pass the translated Unix path to wineserver, which changes directories to `dirname(path)`. It then provides `bind` and `connect` with `basename(path)`. This is not threadsafe, but wineserver is not (currently) multithreaded.
Abstract sockets are not fully supported by this patch, matching the behavior of Windows.
--
v58: ws2_32/tests: Add test for AF_UNIX sockets.
https://gitlab.winehq.org/wine/wine/-/merge_requests/2786
[View Less]
This clears up much of the todo list for mspatcha. GetFilePatchSignature* and NormalizeFileForPatchSignature have now been implemented in these changes.
These changes bring better support for things like the Adobe Acrobat installer without the need for things like the winetricks mspatcha.dll native dll override. Still needed is support for interleaved streams in the LZXD decompression logic, however this is a great start to better supporting software installers that use the Windows interface …
[View More]for creating and applying patches to files.
This is the beginning of the fixes required for bug 12501: https://bugs.winehq.org/show_bug.cgi?id=12501
--
v28: mspatcha: Use string comparison for section names
mspatcha: Use unaligned typedefs
mspatcha: Relocate PE/COFF image functions
mspatcha: Add support for 32-bit file image patch transforms
https://gitlab.winehq.org/wine/wine/-/merge_requests/3870
[View Less]
> Understood. Should the tests be added (`todo_wine`'d) to document the behavior discrepancy or should those only be included with the implementation if it becomes necessary to fix a real application?
If you want to submit the tests alone I think they'd be accepted.
--
https://gitlab.winehq.org/wine/wine/-/merge_requests/3367#note_46211
Looks okay I think although I find the seat locking again a bit overkill. I don't really see a scenario where a wine process would be used across seats, can this really happen on Wayland?
I understand that as the event is dynamically sent we may have to support it, but I would find it simpler if we could assume that we have a seat from the start (eventually waiting for it on startup, idk), and where we'd just abort if the seat is removed.
--
https://gitlab.winehq.org/wine/wine/-/merge_requests/3909#note_46203
This clears up much of the todo list for mspatcha. GetFilePatchSignature* and NormalizeFileForPatchSignature have now been implemented in these changes.
These changes bring better support for things like the Adobe Acrobat installer without the need for things like the winetricks mspatcha.dll native dll override. Still needed is support for interleaved streams in the LZXD decompression logic, however this is a great start to better supporting software installers that use the Windows interface …
[View More]for creating and applying patches to files.
This is the beginning of the fixes required for bug 12501: https://bugs.winehq.org/show_bug.cgi?id=12501
--
v27: mspatcha: Use string comparison for section names
mspatcha: Use unaligned typedefs
mspatcha: Relocate PE/COFF image functions
mspatcha: Add support for 32-bit file image patch transforms
mspatcha: Add implementations for GetFilePatchSignature* routines
https://gitlab.winehq.org/wine/wine/-/merge_requests/3870
[View Less]
Windows 10 [received support](https://devblogs.microsoft.com/commandline/af_unix-comes-to-window… for AF_UNIX sockets in Insider Build 17063. This merge request adds basic support for AF_UNIX sockets to ws2_32 and wineserver.
Of particular note is the difficulty in handling `sun_path`. Most of the functions that allow for translating Windows paths to Unix paths are not accessible from ws2_32. I considered the following options:
* Pass the Windows path to wineserver and do the conversion there.…
[View More]
* This is, as far as I can tell, not possible without major rearchitecting. wineserver does not have functions to translate Windows paths to Unix paths, for obvious reasons.
* Obtain the current working directory of the requesting process and temporarily change directories to there.
* This only handles relative paths and fails for absolute paths, UNC paths, etc.
* Conditionally change directories based on whether the path is relative or not.
* This is error-prone and wineserver does not have the requisite functions to do this cleanly.
I ultimately decided to pass the translated Unix path to wineserver, which changes directories to `dirname(path)`. It then provides `bind` and `connect` with `basename(path)`. This is not threadsafe, but wineserver is not (currently) multithreaded.
Abstract sockets are not fully supported by this patch, matching the behavior of Windows.
--
v56: ws2_32/tests: Add test for AF_UNIX sockets fix.
server: Fix getsockname() and accept() on AF_UNIX sockets.
server: Introduce error when attempting to create a SOCK_DGRAM AF_UNIX socket.
server: Allow for deletion of socket files.
ws2_32: Add support for AF_UNIX sockets.
https://gitlab.winehq.org/wine/wine/-/merge_requests/2786
[View Less]
> You can't assume that the path is null-terminated.
I'm pretty sure I can assume `unix_path` is null-terminated. It's ultimately provided by `wine_nt_to_unix_file_name()` (`wine_get_unix_file_name()` doesn't modify it), with the line:
```c
if (*size > strlen(name)) strcpy( /* unix_path = */ nameA, name );
```
Multiple places across other DLLs assume `wine_get_unix_file_name()` provides a null-terminated string, passing its return value to e.g. `fprintf()` and `strstr()`.
> In fact …
[View More]it shouldn't be, you should use an explicit length instead.
Whether I `strlen()` here or when first getting this buffer from `wine_get_unix_file_name()` seems irrelevant to me; is there something I'm missing?
--
https://gitlab.winehq.org/wine/wine/-/merge_requests/2786#note_46190
[View Less]
This clears up much of the todo list for mspatcha. GetFilePatchSignature* and NormalizeFileForPatchSignature have now been implemented in these changes.
These changes bring better support for things like the Adobe Acrobat installer without the need for things like the winetricks mspatcha.dll native dll override. Still needed is support for interleaved streams in the LZXD decompression logic, however this is a great start to better supporting software installers that use the Windows interface …
[View More]for creating and applying patches to files.
This is the beginning of the fixes required for bug 12501: https://bugs.winehq.org/show_bug.cgi?id=12501
--
v26: mspatcha: Use string comparison for section names
mspatcha: Use unaligned typedefs
mspatcha: Relocate PE/COFF image functions
mspatcha: Add support for 32-bit file image patch transforms
mspatcha: Add implementations for GetFilePatchSignature* routines
mspatcha: Add support for 32-bit file normalization
https://gitlab.winehq.org/wine/wine/-/merge_requests/3870
[View Less]
--
v2: Release 1.9.
vkd3d-shader/d3dbc: Translate sm1 fragment outputs to system values.
vkd3d-shader/dxbc: Map sm4 fragment outputs to system values based on their name.
vkd3d-shader: Add definitions for more fragment output system values.
https://gitlab.winehq.org/wine/vkd3d/-/merge_requests/356
On Thu Sep 21 09:50:20 2023 +0000, Petrichor Park wrote:
> Oh dear. Do I need to worry about the big scary warning at the top
> saying that my merge failed?
> I don't think I did this right.
No, you didn't. Don't stress too much over it, though, git is nice folk, but it's a well established fact that it's not one with whom it's immediately easy to make friends. Hoping to help, here is what I would do now:
```sh
git checkout master
git pull # Ensure that your master is up-…
[View More]to-date
git checkout impl-tangent
git reset --hard b54b3c60 # Recover your last known-good state of this branch (I got the commit hash from the MR history)
git log
git show # Check that's what you intend to submit
git rebase master # Magic is happening here! Basically your commit is replayed as if it was replayed on top of master
# Git will tell you there is a conflict: basically you added line `{"tan", ...},` in `hlsl.y` and, in the same place,
# Nikolay added line `{"tex1D", ...},` a few days ago (commit c5d680d141; how do I know? I used git blame)
# Solving this conflict is simple: both lines have to appear in the file, so edit the file and just remove
# the markers git added around the conflict; but they also have to be in alphabetic order, so perhaps you'll have to swap them
git add -u # This is probably not needed, because git will automatically detect you fixed the conflict; but if it doesn't happen that's how you add the changes to the index
git diff --cached # Check that the diff now looks correct
git rebase --continue # Tell git that the conflict is resolved and it can carry on with the rebase
git log
git show # Check again before force-pushing, one is never too sure
git push -f # The MR is now updated and the scary warning should go away
```
Please feel free to ask if I can help you further.
--
https://gitlab.winehq.org/wine/vkd3d/-/merge_requests/321#note_46176
[View Less]
On Thu Sep 21 09:49:58 2023 +0000, Dallas Strouse wrote:
> Hi, I'm back again :P
> Since it's probably about time for it, and I forgot to send that email a
> while back, I might as well ask now: are there plans to make Wine use
> libdecor at some point in the future, so that there are "native" window
> decorations for desktops that don't provide xdg-decoration? A lot of the
> work here for compositor-initiated window closing is a requirement for
> that.
> There may be …
[View More]performance implications for this, IIRC, since you'd have
> to work with another library's event loop on top of another GUI toolkit
> not made for games, and GTK doesn't support wl_subsurfaces normally
> *iirc*. SDL does it, at least, so there's that. But something to maybe
> consider. I'm satisfied with the default window decorations right now, myself.
Hi! At this point there are no plans to use any external client-side mechanism to draw window decorations. For libdecor, in particular, my impression is that while its abstractions for event dispatching and surface configuration are helpful for typical Wayland clients, they are not a good match for the driver's integration needs, for which we need finer and more direct control.
Some other ideas:
1. Support server-side decorations, which will help with KDE and wlroots based compositors.
2. Create a Wine decoration theme to better match typical GNOME/KDE etc (will only help with apps that don't do their own decorations).
--
https://gitlab.winehq.org/wine/wine/-/merge_requests/3909#note_46175
[View Less]
Windows 10 [received support](https://devblogs.microsoft.com/commandline/af_unix-comes-to-window… for AF_UNIX sockets in Insider Build 17063. This merge request adds basic support for AF_UNIX sockets to ws2_32 and wineserver.
Of particular note is the difficulty in handling `sun_path`. Most of the functions that allow for translating Windows paths to Unix paths are not accessible from ws2_32. I considered the following options:
* Pass the Windows path to wineserver and do the conversion there.…
[View More]
* This is, as far as I can tell, not possible without major rearchitecting. wineserver does not have functions to translate Windows paths to Unix paths, for obvious reasons.
* Obtain the current working directory of the requesting process and temporarily change directories to there.
* This only handles relative paths and fails for absolute paths, UNC paths, etc.
* Conditionally change directories based on whether the path is relative or not.
* This is error-prone and wineserver does not have the requisite functions to do this cleanly.
I ultimately decided to pass the translated Unix path to wineserver, which changes directories to `dirname(path)`. It then provides `bind` and `connect` with `basename(path)`. This is not threadsafe, but wineserver is not (currently) multithreaded.
Abstract sockets are not fully supported by this patch, matching the behavior of Windows.
--
v54: ws2_32/tests: Add test for AF_UNIX sockets fix.
server: Fix getsockname() and accept() on AF_UNIX sockets.
server: Introduce error when attempting to create a SOCK_DGRAM AF_UNIX socket.
server: Allow for deletion of socket files.
ws2_32: Add support for AF_UNIX sockets.
ntdll/unix: Add support for AF_UNIX sockets to multiple functions.
ws2_32: Add afunix.h header.
https://gitlab.winehq.org/wine/wine/-/merge_requests/2786
[View Less]
On Thu Sep 21 08:28:33 2023 +0000, Gabriele wrote:
> I wonder if from this MR, we'll be able to somehow test those changes,
> by using an environment variable or something similar.
> I'd be very interested in testing it out as soon as it gets merged!
An easy and non-intrusive way to experiment with the upstream Wayland driver is:
1. Create the registry key `HKEY_CURRENT_USER\Software\Wine\Drivers\Graphics="x11,wayland"`
2. Unset the `DISPLAY` env. variable, e.g, `DISPLAY= wine ...`, to …
[View More]force the use of the Wayland driver
--
https://gitlab.winehq.org/wine/wine/-/merge_requests/3909#note_46166
[View Less]