December 2020 - Wine-Devel - list.winehq.org

[RFC] Announcing Wayland driver development
by Alexandros Frantzis Dec. 17, 2020

Dec. 17, 2020

Hi all, For some time now I have been working on a Wayland driver for Wine, and it has now reached a good enough state to present it to the wider community. This driver allows users to run Windows GDI and OpenGL applications directly on Wayland compositors without an intermediate layer to translate from X11 to Wayland. This leads to a leaner and more efficient stack. The code currently resides at: https://gitlab.collabora.com/alf/wine/-/commits/wayland Get it with: $ git clone -b wayland https://gitlab.collabora.com/alf/wine/ Configure with (you need the wayland-client and wayland-egl packages): $ ./configure --with-wayland To run, ensure there is a Wayland compositor running, unset DISPLAY (so that Wayland is picked instead of X11) and set WAYLAND_DISPLAY (if needed): $ DISPLAY= WAYLAND_DISPLAY=wayland-0 ./wine ... Here is video showcasing a few Windows applications running with the Wayland driver on the Weston reference compositor: https://youtu.be/br2b8gUy5n8 The Wayland protocol is by design more constrained compared to more traditional display systems like X11. These constraints add some special challenges in the integration of Wayland with Win32. Since Wayland's window model is not based on a single flat 2D co-ordinate space, as X11's was, the Wayland protocol doesn't allow applications to control their absolute position on the screen. For Win32 transient windows (menus, tooltips, etc) the driver tries to work around the lack of absolute positioning by "anchoring" them to an owning Wayland surface and treating them as subsurfaces of that owner. Screen coordinates for such windows are transformed to local coordinates relative to the owning surface, allowing correct placement through relative subsurface movement, which is supported by Wayland. By using heuristics to select the proper owning surface, this approach has led to very good results. Absolute positioning of non-transient top level windows is not supported at this time, and will likely require a (possibly controversial) Wayland extension if it is to ever be pursued. The lack of absolute positioning also has an adverse effect on input handling, when parts of windows reside outside the visible windows display space, but are still full accessible in the Wayland compositor display space (and thus the user). Input events to such areas don't reach their intended coordinates (are clamped to Windows display bounds) leaving the user mystified about why they are unable to interact with a perfectly visible area in their Wayland compositor. The current partial workaround is to force all windows at (0,0) windows display coordinates, to maximize the area which can be interacted with. With this workaround, and as long as windows remain smaller than the display size, input works without issues. Here what's supported at the moment (modulo bugs): * GDI apps, including layered/translucent windows * OpenGL apps * Window resizing * Maximized and fullscreen window states * Mouse and keyboard (only QWERTY) input * Mouse cursors * Menus, tooltips, etc * Single display What needs more work (or is completely missing): * Minimize * Better z-order and activation handling * Keyboard layout support * Grabs * GTK menu mouse handling bug (cannot click to select items) * Multiple displays * Vulkan. Note that that there is another effort at https://github.com/varmd/wine-wayland/ focusing solely on vulkan. My hope is that we will be able to share efforts going forward. Programs I have tried and which are working well (again modulo bugs): * Native Wine/Win32 apps (notepad, regedit etc) * Supertuxkart * 010Editor * Firefox * Stellarium * Battle For Wesnoth * GIMP (except clicking items in GTK menus as noted above) * Unigine Valley My first question to the Wine community is how complete the driver should be before being considered for inclusion. Is the current Wayland driver state enough? If there is interest, I was hoping to get the driver in earlier rather than later, in an experimental capacity, in order to provide a single, official point of development for people wanting to contribute. My second question is about the best way to move forward with upstreaming. The code is currently split between just a few commits, mostly as a result of the experimental nature of this effort, and also because I used existing code from winex11 and wineandroid as my starting point. I would like to split this into smaller commits to make it more review-friendly for the mailing list. I was thinking of splitting by user32 function (and internal dependencies), but since this is going to be an artificial split, I am open to other ideas to make reviewers' lives easier. Looking forward to your feedback and questions! Thanks, Alexandros

5 8

[PATCH 1/2] widl: don't use local methods when writing async interfaces
by Steve Lhomme Dec. 17, 2020

Dec. 17, 2020

The same entries can have a different call_as between the synchronous and asynchronous interfaces and shouldn't be mixed up. --- tools/widl/header.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/tools/widl/header.c b/tools/widl/header.c index 223ab5c5ca9..43be8d74cbe 100644 --- a/tools/widl/header.c +++ b/tools/widl/header.c @@ -1535,7 +1535,7 @@ static void write_com_interface_start(FILE *header, const type_t *iface) fprintf(header,"#define __%s_%sINTERFACE_DEFINED__\n\n", iface->c_name, dispinterface ? "DISP" : ""); } -static void write_com_interface_end(FILE *header, type_t *iface) +static void write_com_interface_end(FILE *header, type_t *iface, int as_async) { int dispinterface = is_attr(iface->attrs, ATTR_DISPINTERFACE); const UUID *uuid = get_attrp(iface->attrs, ATTR_UUID); @@ -1614,7 +1614,8 @@ static void write_com_interface_end(FILE *header, type_t *iface) if (!dispinterface && !winrt_mode) { write_method_proto(header, iface); - write_locals(header, iface, FALSE); + if (!as_async) + write_locals(header, iface, FALSE); fprintf(header, "\n"); } fprintf(header, "#endif /* __%s_%sINTERFACE_DEFINED__ */\n", iface->c_name, dispinterface ? "DISP" : ""); @@ -1808,11 +1809,11 @@ static void write_header_stmts(FILE *header, const statement_list_t *stmts, cons { write_com_interface_start(header, iface); write_header_stmts(header, type_iface_get_stmts(iface), stmt->u.type, TRUE); - write_com_interface_end(header, iface); + write_com_interface_end(header, iface, FALSE); if (async_iface) { write_com_interface_start(header, async_iface); - write_com_interface_end(header, async_iface); + write_com_interface_end(header, async_iface, TRUE); } } else -- 2.29.2

3 9

Re: TestBot job 83503 results: [PATCH] ws2_32: Fix buffer size check in WSAIoctl() for SIO_GET_INTERFACE_LIST.
by Paul Gofman Dec. 17, 2020

Dec. 17, 2020

On 12/16/20 12:10, Marvin wrote: > ws2_32: > sock.c:10534: Test failed: Got unexpected ret 0. > sock.c:10535: Test failed: Got unexpected error 6. > sock.c:10536: Test failed: Got unexpected size 76. Can it be that Wine is not getting updated on Testbot? The tests results are like the changes to test are applied but ws2_32.dll is unchanged.

2 4

[tools] testbot/RestoreVM: Explicitly deleting the old disks may not be necessary.
by Francois Gouget Dec. 17, 2020

Dec. 17, 2020

Normally virsh undefine --remove-all-storage handles it but it does not follow the symbolic links. So tell rm not to complain if it finds no file to delete. Signed-off-by: Francois Gouget <fgouget(a)codeweavers.com> --- testbot/scripts/RestoreVM | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/testbot/scripts/RestoreVM b/testbot/scripts/RestoreVM index a31f3ea71d..bf99e4978a 100755 --- a/testbot/scripts/RestoreVM +++ b/testbot/scripts/RestoreVM @@ -641,7 +641,7 @@ then then fatal_no_cleanup "could not remove the existing '$opt_vm' VM" fi - [ -z "$olddisks" ] || dry_run rm $olddisks + [ -z "$olddisks" ] || dry_run rm -f $olddisks fi -- 2.20.1

1 0

[tools] testbot/TestAgent: Also dereference symbolic links to better find $::RootDir.
by Francois Gouget Dec. 17, 2020

Dec. 17, 2020

This allows putting a symbolic link to TestAgent in $PATH. Signed-off-by: Francois Gouget <fgouget(a)codeweavers.com> --- testbot/scripts/TestAgent | 10 ++++------ 1 file changed, 4 insertions(+), 6 deletions(-) diff --git a/testbot/scripts/TestAgent b/testbot/scripts/TestAgent index 40f443e462..10bdc63d1b 100755 --- a/testbot/scripts/TestAgent +++ b/testbot/scripts/TestAgent @@ -23,12 +23,10 @@ use strict; sub BEGIN { - if ($0 !~ m=^/=) - { - # Turn $0 into an absolute path so it can safely be used in @INC - require Cwd; - $0 = Cwd::cwd() . "/$0"; - } + # Turn $0 into an absolute path so it can safely be used in @INC. + # Also dereference symbolic links to better find $::RootDir. + use Cwd 'abs_path'; + $0 = abs_path($0); if ($0 =~ m=^(/.*)/[^/]+/[^/]+$=) { $::RootDir = $1; -- 2.20.1

1 0

[tools] testbot/RestoreVM: Avoid errors when restoring a VM with no snapshot.
by Francois Gouget Dec. 17, 2020

Dec. 17, 2020

Signed-off-by: Francois Gouget <fgouget(a)codeweavers.com> --- testbot/scripts/RestoreVM | 1 + 1 file changed, 1 insertion(+) diff --git a/testbot/scripts/RestoreVM b/testbot/scripts/RestoreVM index bf99e4978a..a1b9d15846 100755 --- a/testbot/scripts/RestoreVM +++ b/testbot/scripts/RestoreVM @@ -677,6 +677,7 @@ then _dir="$1" for snapxml in "$_dir"/*.xml do + [ -f "$snapxml" ] || continue name=`basename "$snapxml" .xml` ctime=`sed -e 's~^.*<creationTime>$[0-9][0-9]*$</creationTime>.*$~\1~' -e t -e d "$snapxml"` echo "$ctime $name" -- 2.20.1

1 0

[PATCH] ws2_32/tests: Don't fail the test in test_wsaioctl() if only one interface is available.
by Paul Gofman Dec. 17, 2020

Dec. 17, 2020

Fixes test failure on Testbot. Signed-off-by: Paul Gofman <pgofman(a)codeweavers.com> --- dlls/ws2_32/tests/sock.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/dlls/ws2_32/tests/sock.c b/dlls/ws2_32/tests/sock.c index 98cc3881853..2b0a7354f4b 100644 --- a/dlls/ws2_32/tests/sock.c +++ b/dlls/ws2_32/tests/sock.c @@ -10530,7 +10530,7 @@ static void test_wsaioctl(void) ok(size && size != 0xdeadbeef && !(size % sizeof(INTERFACE_INFO)), "Got unexpected size %u.\n", size); size = 0xdeadbeef; - ret = WSAIoctl(s, SIO_GET_INTERFACE_LIST, NULL, 0, buffer, sizeof(INTERFACE_INFO), &size, NULL, NULL); + ret = WSAIoctl(s, SIO_GET_INTERFACE_LIST, NULL, 0, buffer, sizeof(INTERFACE_INFO) - 1, &size, NULL, NULL); ok(ret == -1, "Got unexpected ret %d.\n", ret); ok(WSAGetLastError() == WSAEFAULT, "Got unexpected error %d.\n", WSAGetLastError()); ok(!size, "Got unexpected size %u.\n", size); -- 2.29.2

1 0

[PATCH v2] crypt32/tests: Fix a registerOIDInfo() failure when missing elevated privileges.
by Francois Gouget Dec. 17, 2020

Dec. 17, 2020

Signed-off-by: Francois Gouget <fgouget(a)codeweavers.com> --- With my apologies for the previous botched attempt. dlls/crypt32/tests/oid.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/dlls/crypt32/tests/oid.c b/dlls/crypt32/tests/oid.c index ae03bba90e8..7e7c1310dd2 100644 --- a/dlls/crypt32/tests/oid.c +++ b/dlls/crypt32/tests/oid.c @@ -646,7 +646,8 @@ static void test_registerOIDInfo(void) SetLastError(0xdeadbeef); ret = CryptUnregisterOIDInfo(&info1); ok(!ret, "should fail\n"); - ok(GetLastError() == ERROR_FILE_NOT_FOUND, "got %u\n", GetLastError()); + ok(GetLastError() == ERROR_FILE_NOT_FOUND || + GetLastError() == ERROR_ACCESS_DENIED, "got %u\n", GetLastError()); info2 = CryptFindOIDInfo(CRYPT_OID_INFO_OID_KEY, (void *)test_oid, 0); ok(!info2, "should fail\n"); -- 2.20.1

1 0

[PATCH] winegcc: allow CROSSCC to have a suffix
by Damjan Jovanovic Dec. 17, 2020

Dec. 17, 2020

This allows eg. CROSSCC=i686-w64-mingw32-gcc10. Does not regenerate ./configure. Signed-off-by: Damjan Jovanovic <damjan.jov(a)gmail.com> --- configure.ac | 2 +- tools/winegcc/Makefile.in | 3 ++- tools/winegcc/winegcc.c | 5 ++++- 3 files changed, 7 insertions(+), 3 deletions(-)

2 1

[rebased PATCH v3 1/2] ntdll: Always copy context in call_user_apc_dispatcher() on x64.
by Paul Gofman Dec. 16, 2020

Dec. 16, 2020

Currently, if call_user_apc_dispatcher() is called with nonzero context, there is no guarantee that the provided context is stored above the rsp = context_ptr->Rsp - (sizeof(CONTEXT) + offsetof(frame,ret_addr)) being set. Signed-off-by: Paul Gofman <pgofman(a)codeweavers.com> --- v3: - properly align / reserve stack for setup_user_apc_dispatcher_stack() call. dlls/ntdll/unix/signal_x86_64.c | 62 +++++++++++++++++++++++++-------- 1 file changed, 48 insertions(+), 14 deletions(-) diff --git a/dlls/ntdll/unix/signal_x86_64.c b/dlls/ntdll/unix/signal_x86_64.c index 7181b48728c..3ae3b4c141e 100644 --- a/dlls/ntdll/unix/signal_x86_64.c +++ b/dlls/ntdll/unix/signal_x86_64.c @@ -250,6 +250,22 @@ C_ASSERT((offsetof(struct stack_layout, xstate) == sizeof(struct stack_layout))) C_ASSERT( sizeof(XSTATE) == 0x140 ); C_ASSERT( sizeof(struct stack_layout) == 0x590 ); /* Should match the size in call_user_exception_dispatcher(). */ +/* stack layout when calling an user apc function. + * FIXME: match Windows ABI. */ +struct apc_stack_layout +{ + ULONG64 save_regs[4]; + void *func; + ULONG64 align; + CONTEXT context; + ULONG64 rbp; + ULONG64 rip; +}; + +/* Should match size and offset in call_user_apc_dispatcher(). */ +C_ASSERT( offsetof(struct apc_stack_layout, context) == 0x30 ); +C_ASSERT( sizeof(struct apc_stack_layout) == 0x510 ); + struct syscall_frame { ULONG64 xmm[10 * 2]; /* xmm6-xmm15 */ @@ -266,6 +282,9 @@ struct syscall_frame ULONG64 ret_addr; }; +/* Should match the offset in call_user_apc_dispatcher(). */ +C_ASSERT( offsetof( struct syscall_frame, ret_addr ) == 0xf0); + struct amd64_thread_data { DWORD_PTR dr0; /* 02f0 debug registers */ @@ -2013,31 +2032,46 @@ static void setup_exception( ucontext_t *sigcontext, EXCEPTION_RECORD *rec ) /*********************************************************************** * call_user_apc_dispatcher */ +struct apc_stack_layout * WINAPI setup_user_apc_dispatcher_stack( CONTEXT *context, struct apc_stack_layout *stack ) +{ + CONTEXT c; + + if (!context) + { + c.ContextFlags = CONTEXT_FULL; + NtGetContextThread( GetCurrentThread(), &c ); + context = &c; + } + memmove( &stack->context, context, sizeof(stack->context) ); + return stack; +} + __ASM_GLOBAL_FUNC( call_user_apc_dispatcher, "movq 0x28(%rsp),%rsi\n\t" /* func */ "movq 0x30(%rsp),%rdi\n\t" /* dispatcher */ "movq %gs:0x30,%rbx\n\t" - "jrcxz 1f\n\t" - "movq 0x98(%rcx),%rax\n\t" /* context_ptr->Rsp */ - "leaq -0x5c0(%rax),%rsp\n\t" /* sizeof(CONTEXT) + offsetof(frame,ret_addr) */ - "andq $~15,%rsp\n\t" - "jmp 2f\n" - "1:\tmovq 0x328(%rbx),%rax\n\t" /* amd64_thread_data()->syscall_frame */ - "leaq -0x4d0(%rax),%rsp\n\t" - "andq $~15,%rsp\n\t" "movq %rdx,%r12\n\t" /* ctx */ "movq %r8,%r13\n\t" /* arg1 */ "movq %r9,%r14\n\t" /* arg2 */ - "movq %rsp,%rdx\n\t" /* context */ - "movl $0x10000b,0x30(%rdx)\n\t" /* context.ContextFlags */ - "movq $~1,%rcx\n\t" - "call " __ASM_NAME("NtGetContextThread") "\n\t" - "movq %rsp,%rcx\n\t" /* context */ + "jrcxz 1f\n\t" + "movq 0x98(%rcx),%rdx\n\t" /* context->Rsp */ + "jmp 2f\n\t" + "1:\tmovq 0x328(%rbx),%rax\n\t" /* amd64_thread_data()->syscall_frame */ + "leaq 0xf0(%rax),%rdx\n\t" /* &amd64_thread_data()->syscall_frame->ret_addr */ + "2:\tsubq $0x510,%rdx\n\t" /* sizeof(struct apc_stack_layout) */ + "andq $~0xf,%rdx\n\t" + "addq $8,%rsp\n\t" /* pop return address */ + "cmpq %rsp,%rdx\n\t" + "cmovbq %rdx,%rsp\n\t" + "subq $0x20,%rsp\n\t" + "call " __ASM_NAME("setup_user_apc_dispatcher_stack") "\n\t" + "movq %rax,%rsp\n\t" + "leaq 0x30(%rsp),%rcx\n\t" /* context */ "movq $0xc0,0x78(%rcx)\n\t" /* context.Rax = STATUS_USER_APC */ "movq %r12,%rdx\n\t" /* ctx */ "movq %r13,%r8\n\t" /* arg1 */ "movq %r14,%r9\n" /* arg2 */ - "2:\tmovq $0,0x328(%rbx)\n\t" + "movq $0,0x328(%rbx)\n\t" /* amd64_thread_data()->syscall_frame */ "movq %rsi,0x20(%rsp)\n\t" /* func */ "movq 0xa0(%rcx),%rbp\n\t" /* context.Rbp */ "pushq 0xf8(%rcx)\n\t" /* context.Rip */ -- 2.29.2

1 1