June 2022 - Wine-Devel - list.winehq.org

[PATCH 0/4] MR286: nsiproxy.sys: Add static ARP entries which are always present on Windows.

by Paul Gofman (＠gofman)

Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=53175 Commit 910d58520a6d75cada82d992757ca013099d9345 (iphlpapi: Return ERROR_NO_DATA from GetIpNetTable() if no entries are found.) introduced a regression in Roon application. It looks like the app doesn't tolerate an error return from GetIpNetTable() if there are active interfaces found. The blamed patch still looks correct to me per se, that is, if number of entries ends up being zero Windows returns ERROR_NO_DATA. However, number of entries is never zero on Windows if there is any network adapter. There are two things why Windows always has some arp entries and we do not: 1. We currently parse /proc/net/arp line on Linux a bit wrong and end up getting no entries from there (patch 1 fixes that); 2. With the patch patch 1 there Roon seems happy (at least as far as I could test it right at start), at least as long we have an entry in /proc/sys/net (which might not be necessarily the case if there is, e. g., no actual network connection). But that breaks SCP: Secret Laboratory again the same way as before 910d58520a6d75cada82d992757ca013099d9345. It turns out the latter game depends on GetIpNetTable() returning at least 4 entries when returning success (for some reason, I am not sure why but I think I saw that quite clearly). And both apps can actually work reliably on Windows as there are always at least 4 static arp entries if there is any network adaptere configured at all. There are "224.0.0.22" and "239.255.255.250" static arp multicast entries which are always there for any interface on Windows, including loopback (starting from Win10; earlier versions might be missing "239.255.255.250" but have "224.0.0.22"). So adding those like on Windows should hopefully fix all those cases. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/286

June 21, 2022

Re: [PATCH v3 0/2] MR283: ntdll: Filter EFlags from user-supplied CONTEXT on x86 and x64.

by Jinoh Kang (＠iamahuman)

Clarified the "atomic" wording in the comment that explains the rationale for SignalObjectAndWait. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/283#note_2415

June 21, 2022

[PATCH v3 2/2] ntdll: Filter EFlags from user-supplied CONTEXT on x86 and x64.

by Jinoh Kang

From: Jinoh Kang <jinoh.kang.kr(a)gmail.com> Signed-off-by: Jinoh Kang <jinoh.kang.kr(a)gmail.com> --- dlls/ntdll/tests/exception.c | 2 -- dlls/ntdll/unix/signal_i386.c | 2 +- dlls/ntdll/unix/signal_x86_64.c | 4 ++-- dlls/ntdll/unix/thread.c | 20 +++++++++++++++++++- dlls/ntdll/unix/unix_private.h | 17 +++++++++++++++++ 5 files changed, 39 insertions(+), 6 deletions(-) diff --git a/dlls/ntdll/tests/exception.c b/dlls/ntdll/tests/exception.c index 73383a09f17..5101839230b 100644 --- a/dlls/ntdll/tests/exception.c +++ b/dlls/ntdll/tests/exception.c @@ -10725,7 +10725,6 @@ static void test_eflags_sanitization(void) ok(status == STATUS_SUCCESS, "NtSetContextThread failed with %lx\n", status); status = pNtGetContextThread(thread, &context); ok(status == STATUS_SUCCESS, "NtGetContextThread failed with %lx\n", status); - todo_wine ok(context.EFlags == min_eflags, "Expected EFlags to be %#lx, got %#lx\n", min_eflags, context.EFlags); context.EFlags = -1; @@ -10733,7 +10732,6 @@ static void test_eflags_sanitization(void) ok(status == STATUS_SUCCESS, "NtSetContextThread failed with %lx\n", status); status = pNtGetContextThread(thread, &context); ok(status == STATUS_SUCCESS, "NtGetContextThread failed with %lx\n", status); - todo_wine ok(context.EFlags == max_eflags || broken(context.EFlags == max_eflags_broken), "Expected EFlags to be %#lx, got %#lx\n", max_eflags, context.EFlags); diff --git a/dlls/ntdll/unix/signal_i386.c b/dlls/ntdll/unix/signal_i386.c index 7be0c39c424..9be0769904d 100644 --- a/dlls/ntdll/unix/signal_i386.c +++ b/dlls/ntdll/unix/signal_i386.c @@ -941,7 +941,7 @@ NTSTATUS WINAPI NtSetContextThread( HANDLE handle, const CONTEXT *context ) frame->esp = context->Esp; frame->ebp = context->Ebp; frame->eip = context->Eip; - frame->eflags = context->EFlags; + frame->eflags = arch_flags_reg_from_user( context->EFlags, IMAGE_FILE_MACHINE_I386 ); frame->cs = context->SegCs; frame->ss = context->SegSs; } diff --git a/dlls/ntdll/unix/signal_x86_64.c b/dlls/ntdll/unix/signal_x86_64.c index 6c87e347eac..c4f67a16cc9 100644 --- a/dlls/ntdll/unix/signal_x86_64.c +++ b/dlls/ntdll/unix/signal_x86_64.c @@ -1788,7 +1788,7 @@ NTSTATUS WINAPI NtSetContextThread( HANDLE handle, const CONTEXT *context ) frame->rsp = context->Rsp; frame->rbp = context->Rbp; frame->rip = context->Rip; - frame->eflags = context->EFlags; + frame->eflags = arch_flags_reg_from_user( context->EFlags, IMAGE_FILE_MACHINE_AMD64 ); frame->cs = context->SegCs; frame->ss = context->SegSs; } @@ -2002,7 +2002,7 @@ NTSTATUS set_thread_wow64_context( HANDLE handle, const void *ctx, ULONG size ) wow_frame->Esp = context->Esp; wow_frame->Ebp = context->Ebp; wow_frame->Eip = context->Eip; - wow_frame->EFlags = context->EFlags; + wow_frame->EFlags = arch_flags_reg_from_user( context->EFlags, IMAGE_FILE_MACHINE_I386 ); wow_frame->SegCs = cs32_sel; wow_frame->SegSs = ds64_sel; cpu->Flags |= WOW64_CPURESERVED_FLAG_RESET_STATE; diff --git a/dlls/ntdll/unix/thread.c b/dlls/ntdll/unix/thread.c index ad47a5fce74..e0d44bb4a38 100644 --- a/dlls/ntdll/unix/thread.c +++ b/dlls/ntdll/unix/thread.c @@ -1663,13 +1663,31 @@ NTSTATUS WINAPI NtQueueApcThread( HANDLE handle, PNTAPCFUNC func, ULONG_PTR arg1 NTSTATUS set_thread_context( HANDLE handle, const void *context, BOOL *self, USHORT machine ) { context_t server_contexts[2]; - unsigned int count = 0; + unsigned int count = 0, i; NTSTATUS ret; context_to_server( &server_contexts[count++], native_machine, context, machine ); if (machine != native_machine) context_to_server( &server_contexts[count++], machine, context, machine ); + for (i = 0; i < count; i++) + { + unsigned int *flags_ptr; + if (!(server_contexts[i].flags & SERVER_CTX_CONTROL)) continue; + switch (server_contexts[i].machine) + { + case IMAGE_FILE_MACHINE_I386: + flags_ptr = &server_contexts[i].ctl.i386_regs.eflags; + break; + case IMAGE_FILE_MACHINE_AMD64: + flags_ptr = &server_contexts[i].ctl.x86_64_regs.flags; + break; + default: + continue; + } + *flags_ptr = arch_flags_reg_from_user( *flags_ptr, machine ); + } + SERVER_START_REQ( set_thread_context ) { req->handle = wine_server_obj_handle( handle ); diff --git a/dlls/ntdll/unix/unix_private.h b/dlls/ntdll/unix/unix_private.h index 795fc148479..4160f731563 100644 --- a/dlls/ntdll/unix/unix_private.h +++ b/dlls/ntdll/unix/unix_private.h @@ -458,4 +458,21 @@ static inline NTSTATUS map_section( HANDLE mapping, void **ptr, SIZE_T *size, UL 0, NULL, size, ViewShare, 0, protect ); } +static inline DWORD arch_flags_reg_from_user( DWORD flags, USHORT machine ) +{ + switch (machine) + { + case IMAGE_FILE_MACHINE_I386: + if (machine == native_machine) + { + return (flags & 0x003f4fd7) | 0x00000200; + } + return (flags & 0x003f0fd7) | 0x00000202; + case IMAGE_FILE_MACHINE_AMD64: + return (flags & 0x00210fd5) | 0x00000200; + default: + return flags; + } +} + #endif /* __NTDLL_UNIX_PRIVATE_H */ -- GitLab https://gitlab.winehq.org/wine/wine/-/merge_requests/283

June 21, 2022

[PATCH v3 1/2] ntdll/tests: Test for setting x86 / x64 EFLAGS register in context.

by Jinoh Kang

From: Jinoh Kang <jinoh.kang.kr(a)gmail.com> Signed-off-by: Jinoh Kang <jinoh.kang.kr(a)gmail.com> --- dlls/ntdll/tests/exception.c | 91 ++++++++++++++++++++++++++++++++++++ 1 file changed, 91 insertions(+) diff --git a/dlls/ntdll/tests/exception.c b/dlls/ntdll/tests/exception.c index 820e435bc1b..73383a09f17 100644 --- a/dlls/ntdll/tests/exception.c +++ b/dlls/ntdll/tests/exception.c @@ -10673,6 +10673,95 @@ static void test_copy_context(void) dst_xs->Mask == 4, sizeof(dst_xs->YmmContext)); } } + +static DWORD WINAPI signal_and_wait_proc( void *arg ) +{ + HANDLE *events = arg; + /* Signal event and block thread without returning to userspace. */ + return SignalObjectAndWait(events[0], events[1], INFINITE, FALSE); +} + +static void test_eflags_sanitization(void) +{ + DWORD min_eflags, max_eflags, max_eflags_broken; + HANDLE event_in, event_out; + HANDLE thread; + HANDLE thread_args[2]; + HANDLE waits[2]; + NTSTATUS status; + CONTEXT context; + DWORD result, old_eflags; + int i; + +#if defined(__x86_64__) + min_eflags = 0x00000200; + max_eflags = 0x00210fd5; + max_eflags_broken = max_eflags; +#elif defined(__i386__) + min_eflags = is_wow64 ? 0x00000202 : 0x00000200; + max_eflags = is_wow64 ? 0x003f0fd7 : 0x003f4fd7; + max_eflags_broken = max_eflags & ~((DWORD)!is_wow64 << 17); +#endif + + event_in = CreateEventW(NULL, FALSE, FALSE, NULL); + ok(event_in != NULL, "CreateEventW failed with %#lx\n", GetLastError()); + event_out = CreateEventW(NULL, FALSE, FALSE, NULL); + ok(event_out != NULL, "CreateEventW failed with %#lx\n", GetLastError()); + + thread_args[0] = event_in; + thread_args[1] = event_out; + thread = CreateThread(NULL, 0, signal_and_wait_proc, thread_args, CREATE_SUSPENDED, NULL); + ok(thread != NULL, "CreateThread failed with %#lx.\n", GetLastError()); + + for (i = 0; ; i++) + { + context.ContextFlags = CONTEXT_CONTROL; + status = pNtGetContextThread(thread, &context); + ok(status == STATUS_SUCCESS, "NtGetContextThread failed with %lx\n", status); + old_eflags = context.EFlags; + + context.EFlags = 0; + status = pNtSetContextThread(thread, &context); + ok(status == STATUS_SUCCESS, "NtSetContextThread failed with %lx\n", status); + status = pNtGetContextThread(thread, &context); + ok(status == STATUS_SUCCESS, "NtGetContextThread failed with %lx\n", status); + todo_wine + ok(context.EFlags == min_eflags, "Expected EFlags to be %#lx, got %#lx\n", min_eflags, context.EFlags); + + context.EFlags = -1; + status = pNtSetContextThread(thread, &context); + ok(status == STATUS_SUCCESS, "NtSetContextThread failed with %lx\n", status); + status = pNtGetContextThread(thread, &context); + ok(status == STATUS_SUCCESS, "NtGetContextThread failed with %lx\n", status); + todo_wine + ok(context.EFlags == max_eflags || broken(context.EFlags == max_eflags_broken), + "Expected EFlags to be %#lx, got %#lx\n", max_eflags, context.EFlags); + + context.EFlags = old_eflags; + status = pNtSetContextThread(thread, &context); + ok(status == STATUS_SUCCESS, "NtSetContextThread failed with %lx\n", status); + + if (i == 1) break; + + ResumeThread(thread); + waits[0] = event_in; + waits[1] = thread; + result = WaitForMultipleObjects(2, waits, FALSE, INFINITE); + ok(result == WAIT_OBJECT_0, "Expected WAIT_OBJECT_0, got %#lx\n", result); + } + + ok(SetEvent(event_out), "SetEvent failed: %#lx\n", GetLastError()); + + result = WaitForSingleObject(thread, INFINITE); + ok(result == WAIT_OBJECT_0, "Expected WAIT_OBJECT_0, got %#lx\n", result); + + ok(GetExitCodeThread(thread, &result), "Failed to get thread exit code: %#lx\n", GetLastError()); + ok(result == WAIT_OBJECT_0, "Expected WAIT_OBJECT_0, got %#lx\n", result); + + CloseHandle(thread); + CloseHandle(event_in); + CloseHandle(event_out); +} #endif START_TEST(exception) @@ -10847,6 +10936,7 @@ START_TEST(exception) test_kiuserexceptiondispatcher(); test_extended_context(); test_copy_context(); + test_eflags_sanitization(); #elif defined(__x86_64__) @@ -10891,6 +10981,7 @@ START_TEST(exception) test_copy_context(); test_unwind_from_apc(); test_syscall_clobbered_regs(); + test_eflags_sanitization(); #elif defined(__aarch64__) -- GitLab https://gitlab.winehq.org/wine/wine/-/merge_requests/283

June 21, 2022

[PATCH v3 0/2] MR283: ntdll: Filter EFlags from user-supplied CONTEXT on x86 and x64.

by Jinoh Kang (＠iamahuman)

-- v3: ntdll: Filter EFlags from user-supplied CONTEXT on x86 and x64. ntdll/tests: Test for setting x86 / x64 EFLAGS register in context. https://gitlab.winehq.org/wine/wine/-/merge_requests/283

June 21, 2022

Re: [PATCH 1/2] ntdll/tests: Test for setting x86 / x64 EFLAGS register in context.

by Jinoh Kang

On 6/21/22 00:34, Francois Gouget wrote: > On Mon, 20 Jun 2022, Marvin wrote: > [...] >> === w7u_2qxl (32 bit report) === >> === w8 (32 bit report) === >> >> ntdll: >> exception.c:10725: Test failed: Expected EFlags to be 0x3f0fd7, got 0x3f4fd7 >> exception.c:10733: Test failed: Expected EFlags to be 0x3f0fd7, got 0x200 >> exception.c:10725: Test failed: Expected EFlags to be 0x3f0fd7, got 0x3f4fd7 >> exception.c:10733: Test failed: Expected EFlags to be 0x3f0fd7, got 0x200 > > As far as I can tell these are indeed new failures. So it looks like > this patch needs to be tweaked to work on Windows 7 and 8. I have resolved that in v2. -- Sincerely, Jinoh Kang

June 21, 2022

Re: [PATCH 0/1] MR132: opengl32: Don't prioritize low bit depth formats with non-matching stencil.

by Matteo Bruni (＠Mystral)

On Mon Jun 20 22:00:21 2022 +0000, Matteo Bruni wrote: > It seems pretty clear and reasonable to me actually. It looks like > ChoosePixelFormat() returns a pixel format with stencil bits when > requested so. Matching depth bits comes at a lower priority. > As it turns out, the only actual pixel format with stencil bits that's > supported everywhere is D24S8. D32S8 is apparently supported on Windows > AMD, but not on Windows Nvidia. FWIW Linux AMD doesn't return any D32 > visual / fbconfig at all. > I'm attaching some more changes on top of yours: > [choosepixelformat.txt](/uploads/b892cfcd3cb9fb13f8eee3ab81b2ddfb/choosepixelformat.txt). > Basically I swapped the two blocks for stencil and depth in > wglChoosePixelFormat(), making sure we take care of stencil before > checking depth. Also adding the output of the tests on Windows Nvidia after my patch. [output_nvidia.txt](/uploads/99fd7148e6a77bfea79713e71aa0a03b/output_nvidia.txt) -- https://gitlab.winehq.org/wine/wine/-/merge_requests/132#note_2411

June 20, 2022

Re: [PATCH 0/1] MR132: opengl32: Don't prioritize low bit depth formats with non-matching stencil.

by Matteo Bruni (＠Mystral)

On Mon Jun 20 16:08:35 2022 +0000, Paul Gofman wrote: > I run the test on my AMD / Windows and added some traces / additional > tests. I am attaching diff to the test (which includes your patch as > well) and the output from AMD / Windows. > It seems that unfortunately 32/8 tests is a bit inconclusive here as > somehow the output pixel format is 32 / 8 (honestly not sure what that > means but that's what I see here; see no test failure on line 382 and > trace output from line 381). > The rests of tests suggest that it prefers 24 bit whenever in doubt, > see, e. g., 8x8 test, trace at line 364: it could choose 16x8 but > preferred 24x8. From what I see it seems that the pattern is whenever > stencil is requested it returns depths >= 24. That's what my current > patch is doing. If we prioritize stencil formats when requested over > depth match that will probably look more straightforward logic-wise but > that would break all those tests if we'd tighten them to what is > actually returned on AMD. Also, if specific games depending on the > stencil choice are concerned, if we plainly prioritize stencil presence > they will get 16 bit depth on Wine while getting 24 on Windows and that > difference may matter (even if not break the thing completely like > returning no stencil format). > What do you think? > [diff.txt](/uploads/4355976f66b6decb44d43c5b126e5479/diff.txt) > [output.txt](/uploads/13bfe7a03d34c1303f04612c9d809e0e/output.txt) It seems pretty clear and reasonable to me actually. It looks like ChoosePixelFormat() returns a pixel format with stencil bits when requested so. Matching depth bits comes at a lower priority. As it turns out, the only actual pixel format with stencil bits that's supported everywhere is D24S8. D32S8 is apparently supported on Windows AMD, but not on Windows Nvidia. FWIW Linux AMD doesn't return any D32 visual / fbconfig at all. I'm attaching some more changes on top of yours: [choosepixelformat.txt](/uploads/b892cfcd3cb9fb13f8eee3ab81b2ddfb/choosepixelformat.txt). Basically I swapped the two blocks for stencil and depth in wglChoosePixelFormat(), making sure we take care of stencil before checking depth. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/132#note_2409

June 20, 2022

Re: [PATCH v3 0/1] MR259: d3d11: Remove null dxgi object checks. - approved

by Alexandre Julliard (＠julliard)

This merge request was approved by Alexandre Julliard. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/259

June 20, 2022

Re: [PATCH v3 0/1] MR279: d3drm: IDirect3DRMMeshBuilder3 Optimize fake success - approved

by Alexandre Julliard (＠julliard)

This merge request was approved by Alexandre Julliard. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/279

June 20, 2022