October 2017 - Wine-devel - list.winehq.org

Re: [PATCH] msvcp120: Fix _Xtime_diff_to_millis2 overflow behavior.
by Piotr Caban 17 Oct '17

17 Oct '17

On 10/11/17 17:07, Stefan Dösinger wrote: > + {0x7FFFFFFFFFFFFFFF, 0, 0x7FFFFFFFFFFFFFFD, 0, 0}, /* Not an overflow */ Defining LONGLONG constants this way is not portable. > @@ -410,7 +410,13 @@ MSVCRT_long __cdecl _Xtime_diff_to_millis2(const xtime *t1, const xtime *t2) > diff_nsec = t1->nsec - t2->nsec; > ret = diff_sec * MILLISEC_PER_SEC + > (diff_nsec + NANOSEC_PER_MILLISEC - 1) / NANOSEC_PER_MILLISEC; > - return ret > 0 ? ret : 0; > + > + /* This function does not return negative numbers, except when they get created through > + * an integer overflow. */ > + if (diff_sec > ((LONGLONG)INT_MAX + NANOSEC_PER_MILLISEC - 1) / NANOSEC_PER_MILLISEC) > + return ret; > + else > + return ret > 0 ? ret : 0; > } It looks like the function should be treating negative time differences in special way. Except of that it seems to be ignoring overflows. I think that following code is much easier to read. Is something like this working for you? MSVCRT_long __cdecl _Xtime_diff_to_millis2(const xtime *t1, const xtime *t2) { LONGLONG diff_sec, diff_nsec; TRACE("(%p, %p)\n", t1, t2); diff_sec = t1->sec - t2->sec; diff_nsec = t1->nsec - t2->nsec; diff_sec += diff_nsec / NANOSEC_PER_SEC; diff_nsec %= NANOSEC_PER_SEC; if (diff_nsec < 0) { diff_sec -= 1; diff_nsec += NANOSEC_PER_SEC; } if (diff_sec<0 || (diff_sec==0 && diff_nsec<0)) return 0; return diff_sec * MILLISEC_PER_SEC + (diff_nsec + NANOSEC_PER_MILLISEC - 1) / NANOSEC_PER_MILLISEC; } Thanks, Piotr

2 4

Re: [PATCH] ntdll: Always restore TEB to x18 on aarch 64 on return from calls to builtins
by Stefan Dösinger 17 Oct '17

17 Oct '17

Am 2017-10-13 um 09:52 schrieb Martin Storsjo: > This requires always enabling relaying of calls though. It also won't work for COM object calls, which aren't intercepted by +relay.

1 0

Re: [PATCH 1/2] mshtml: Use window event target instead of body only for selected event properties.
by Marvin 17 Oct '17

17 Oct '17

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=33544 Your paranoid android. === w7u (32 bit htmldoc) === htmldoc.c:2682: Test failed: unexpected call GetOverrideKeyPath

1 0

Re: [PATCH] Revert "user32/combo: Set listbox popup height correctly and add tests."
by Marvin 16 Oct '17

16 Oct '17

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=33545 Your paranoid android. === build (build) === The test VM is powered off!

1 0

Re: [PATCH] winex11.drv: Fix misplaced parentheses.
by Alexandre Julliard 16 Oct '17

16 Oct '17

Andrey Gusev <andrey.goosev(a)gmail.com> writes: > Signed-off-by: Andrey Gusev <andrey.goosev(a)gmail.com> > --- > dlls/winex11.drv/mouse.c | 2 +- > dlls/winex11.drv/systray.c | 2 +- > dlls/winex11.drv/window.c | 2 +- > 3 files changed, 3 insertions(+), 3 deletions(-) Not that it makes much difference, but the existing code is correct. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: [PATCH] regedit: Correctly import hex values with no data
by Marvin 16 Oct '17

16 Oct '17

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=33540 Your paranoid android. === build (build) === Patch failed to apply

2 1

Re: [PATCH] reg/tests: Correctly import hex values with no data
by Marvin 16 Oct '17

16 Oct '17

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=33541 Your paranoid android. === build (build) === Patch failed to apply

2 1

Re: [01/10] secur32: Implement AcquireCredentialsHandle for Kerberos.
by Alexandre Julliard 16 Oct '17

16 Oct '17

Hans Leidekker <hans(a)codeweavers.com> writes: > @@ -2,8 +2,8 @@ MODULE = secur32.dll > IMPORTLIB = secur32 > IMPORTS = netapi32 advapi32 > DELAYIMPORTS = crypt32 > -EXTRAINCL = $(GNUTLS_CFLAGS) > -EXTRALIBS = $(SECURITY_LIBS) > +EXTRAINCL = $(GNUTLS_CFLAGS) $(GSSAPI_CFLAGS) > +EXTRALIBS = $(SECURITY_LIBS) $(GSSAPI_LIBS) Shouldn't the library be loaded dynamically? -- Alexandre Julliard julliard(a)winehq.org

2 2

Re: secur32/tests: Add tests for the Kerberos provider.
by Marvin 16 Oct '17

16 Oct '17

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=33537 Your paranoid android. === wxppro (32 bit kerberos) === kerberos.c:361: Test failed: got 00073bbf kerberos.c:149: Test failed: AcquireCredentialsHandleA returned 80090311 === w2003std (32 bit kerberos) === kerberos.c:361: Test failed: got 00073bbf kerberos.c:149: Test failed: AcquireCredentialsHandleA returned 8009030e === wvistau64 (32 bit kerberos) === kerberos.c:149: Test failed: AcquireCredentialsHandleA returned 8009030e === w2008s64 (32 bit kerberos) === kerberos.c:149: Test failed: AcquireCredentialsHandleA returned 8009030e === w7u (32 bit kerberos) === kerberos.c:149: Test failed: AcquireCredentialsHandleA returned 8009030e === w7pro64 (32 bit kerberos) === kerberos.c:149: Test failed: AcquireCredentialsHandleA returned 8009030e === w8 (32 bit kerberos) === kerberos.c:361: Test failed: got 008f3bbf kerberos.c:149: Test failed: AcquireCredentialsHandleA returned 8009030e === w864 (32 bit kerberos) === kerberos.c:361: Test failed: got 008f3bbf kerberos.c:149: Test failed: AcquireCredentialsHandleA returned 8009030e === w1064 (32 bit kerberos) === kerberos.c:361: Test failed: got 008f3bbf kerberos.c:149: Test failed: AcquireCredentialsHandleA returned 8009030e === wvistau64 (64 bit kerberos) === kerberos.c:149: Test failed: AcquireCredentialsHandleA returned 8009030e === w2008s64 (64 bit kerberos) === kerberos.c:149: Test failed: AcquireCredentialsHandleA returned 8009030e === w7pro64 (64 bit kerberos) === kerberos.c:149: Test failed: AcquireCredentialsHandleA returned 8009030e === w864 (64 bit kerberos) === kerberos.c:361: Test failed: got 008f3bbf kerberos.c:149: Test failed: AcquireCredentialsHandleA returned 8009030e === w1064 (64 bit kerberos) === kerberos.c:361: Test failed: got 008f3bbf kerberos.c:149: Test failed: AcquireCredentialsHandleA returned 8009030e

1 0

Re: [PATCH] comctl32/tests: Fix treeview test failures on French Windows XP.
by Nikolay Sivov 16 Oct '17

16 Oct '17

On 16.10.2017 6:33, Alex Henrie wrote: > The one testbot machine that's having trouble with these is the one XP > machine with a French locale: > http://test.winehq.org/data/de7220e2553764afd5cf8d463110d4ec5b1a6696/index_… > > Signed-off-by: Alex Henrie <alexhenrie24(a)gmail.com> > --- > dlls/comctl32/tests/treeview.c | 7 +++++++ > 1 file changed, 7 insertions(+) > > diff --git a/dlls/comctl32/tests/treeview.c b/dlls/comctl32/tests/treeview.c > index e7ce92cae8..26ba160c60 100644 > --- a/dlls/comctl32/tests/treeview.c > +++ b/dlls/comctl32/tests/treeview.c > @@ -220,6 +220,10 @@ static const struct message test_right_click_seq[] = { > { WM_RBUTTONDOWN, sent|wparam, MK_RBUTTON }, > { WM_CAPTURECHANGED, sent|defwinproc }, > { TVM_GETNEXTITEM, sent|wparam|lparam|defwinproc, TVGN_CARET, 0 }, > + /* the optional messages below are for French WinXP */ > + { WM_NCHITTEST, sent|optional }, > + { WM_SETCURSOR, sent|optional }, > + { WM_MOUSEFIRST, sent|optional }, > { 0 } > }; > > @@ -367,6 +371,9 @@ static const struct message parent_vk_return_seq[] = { > static const struct message parent_right_click_seq[] = { > { WM_NOTIFY, sent|id, 0, 0, NM_RCLICK }, > { WM_CONTEXTMENU, sent }, > + /* the optional messages below are for French WinXP */ > + { WM_NOTIFY, sent|optional }, > + { WM_SETCURSOR, sent|optional }, > { 0 } > }; > If it works, that's great, but comment makes no sense. I find it hard to believe UI language affects messages sent to windows, must be something else. Please resend without it.

1 0