October 2006 - Wine-Devel - list.winehq.org

Re: Scrambled VB Fonts improving.

by Tom Spear

Joe, not to sound rude, or anything, but please send emails as plain text to the wine list. Any pictures you need to show us, please attach them, and just name them appropriately. The reason we request this is due to different client's handling (or lack thereof) of HTML. It makes the message difficult to read, regardless though. Thanks! Tom On 9/28/06, Joe Baker <joebaker(a)dcresearch.com> wrote: > > I've been having problems with Visual Basic applications under Wine at > first under VNC, then more lately, also when running them against X servers > from the LTSP 4.2 project. > > Here's some illustrations of how bad the font rendering was looking as run > against a VNC server and how much improvement there has recently been. > -snip- -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://www.imvu.com/catalog/web_registration.php?userId=1547373

Oct. 2, 2006

Request for Test: exception handling test

by Marcus Meissner

Hi, Can someone with ability to run the testsuite on Windows please test the following new 2 testcases under Windows? They succeed in WINE now, but I wonder if they will work in Windows. Ciao, Marcus --- dlls/ntdll/tests/exception.c | 145 ++++++++++++++++++++++++++++++++++++++++++ 1 files changed, 145 insertions(+), 0 deletions(-) ef86b7cd65ade45a1e1cd269bff378166d1ad677 diff --git a/dlls/ntdll/tests/exception.c b/dlls/ntdll/tests/exception.c index 1d00816..20ff896 100644 --- a/dlls/ntdll/tests/exception.c +++ b/dlls/ntdll/tests/exception.c @@ -24,6 +24,7 @@ #define _WIN32_WINNT 0x500 /* For NTSTATUS */ #endif +#include <stdio.h> #include "ntstatus.h" #define WIN32_NO_STATUS #include "windef.h" @@ -281,6 +282,148 @@ static void test_debug_regs(void) ok(ctx.Dr7 == 0x155,"failed to set debugregister 7 to 0x155"); } +/* test the single step exception behaviour */ +static int gotsinglesteps = 0; +static DWORD single_step_handler( EXCEPTION_RECORD *rec, EXCEPTION_REGISTRATION_RECORD *frame, + CONTEXT *context, EXCEPTION_REGISTRATION_RECORD **dispatcher ) +{ + gotsinglesteps++; + ok (!(context->EFlags & 0x100), "eflags has single stepping bit set\n"); + return ExceptionContinueExecution; +} + +static BYTE single_stepcode[] = { + 0x9c, /* pushf */ + 0x58, /* pop %eax */ + 0x0d,0,1,0,0, /* or $0x100,%eax */ + 0x50, /* push %eax */ + 0x9d, /* popf */ + 0x35,0,1,0,0, /* xor $0x100,%eax */ + 0x50, /* push %eax */ + 0x9d, /* popf */ + 0xc3 +}; + +static void test_single_step(void) +{ + struct { + EXCEPTION_REGISTRATION_RECORD frame; + } exc_frame; + void (*func)(void) = (void *)single_stepcode; + + pNtCurrentTeb = (void *)GetProcAddress( GetModuleHandleA("ntdll.dll"), "NtCurrentTeb" ); + if (!pNtCurrentTeb) { + trace( "NtCurrentTeb not found, skipping tests\n" ); + return; + } + exc_frame.frame.Handler = single_step_handler; + exc_frame.frame.Prev = pNtCurrentTeb()->Tib.ExceptionList; + pNtCurrentTeb()->Tib.ExceptionList = &exc_frame.frame; + func(); + ok(gotsinglesteps == 1, "expected 1 single step exceptions, got %d\n", gotsinglesteps); + pNtCurrentTeb()->Tib.ExceptionList = exc_frame.frame.Prev; +} + +/* Test the alignment check (AC) flag handling. */ +static BYTE align_check_code[] = { + 0x55, /* push %ebp */ + 0x89,0xe5, /* mov %esp,%ebp */ + 0x9c, /* pushf */ + 0x58, /* pop %eax */ + 0x0d,0,0,4,0, /* or $0x40000,%eax */ + 0x50, /* push %eax */ + 0x9d, /* popf */ + 0x8b,0x45,8, /* mov 0x8(%ebp),%eax */ + 0xc7,0x40,1,42,0,0,0, /* movl $42,0x1(%eax) */ + 0x9c, /* pushf */ + 0x58, /* pop %eax */ + 0x35,0,0,4,0, /* xor $0x40000,%eax */ + 0x50, /* push %eax */ + 0x9d, /* popf */ + 0x5d, /* pop %ebp */ + 0xc3, /* ret */ +}; + +static int gotalignfaults = 0; +static DWORD align_check_handler( EXCEPTION_RECORD *rec, EXCEPTION_REGISTRATION_RECORD *frame, + CONTEXT *context, EXCEPTION_REGISTRATION_RECORD **dispatcher ) +{ + ok (!(context->EFlags & 0x40000), "eflags has AC bit set\n"); + gotalignfaults++; + return ExceptionContinueExecution; +} + +/* Not working in Linux now. Causes problems with unrelated functions. + * have to fix them sometime :( + */ +#undef STACK_ALIGN_TEST + +#ifdef STACK_ALIGN_TEST +static BYTE align_check_code2[] = { + 0x55, /* pushl %ebp */ + 0x89,0xe5, /* movl %esp,%ebp */ + 0x9c, /* pushf */ + 0x58, /* pop %eax */ + 0x0d,0,0,4,0, /* or $0x40000,%eax */ + 0x50, /* push %eax */ + 0x9d, /* popf */ + 0x35,0,0,4,0, /* xor $0x40000,%eax */ + 0x83,0xec,2, /* sub $0x2,%esp */ + 0x50, /* push %eax - test fault on stack */ + 0x89,0xec, /* movl %ebp, %esp */ + 0x50, /* push %eax */ + 0x9d, /* popf */ + 0x5d, /* popl %ebp */ + 0xcb /* lret */ +}; + +static DWORD align_check_handler2( EXCEPTION_RECORD *rec, EXCEPTION_REGISTRATION_RECORD *frame, + CONTEXT *context, EXCEPTION_REGISTRATION_RECORD **dispatcher ) +{ + ok (!(context->EFlags & 0x40000), "eflags has AC bit set\n"); + if (context->Esp & 0x2) { + gotalignfaults++; + context->Esp+=2; + } + return ExceptionContinueExecution; +} +#endif + + +static void test_align_faults(void) +{ + int twoints[2]; + struct { + EXCEPTION_REGISTRATION_RECORD frame; + } exc_frame; + void (*func1)(int*) = (void *)align_check_code; +#ifdef STACK_ALIGN_TEST + void (*func2)(void) = (void *)align_check_code2; +#endif + + pNtCurrentTeb = (void *)GetProcAddress( GetModuleHandleA("ntdll.dll"), "NtCurrentTeb" ); + if (!pNtCurrentTeb) { + trace( "NtCurrentTeb not found, skipping tests\n" ); + return; + } + exc_frame.frame.Handler = align_check_handler; + exc_frame.frame.Prev = pNtCurrentTeb()->Tib.ExceptionList; + pNtCurrentTeb()->Tib.ExceptionList = &exc_frame.frame; + + gotalignfaults=0; + func1(twoints); + ok(gotalignfaults == 0, "got %d alignment faults, expected 0\n", gotalignfaults); + +#ifdef STACK_ALIGN_TEST + exc_frame.frame.Handler = align_check_handler2; + gotalignfaults=0; + func2(); + ok(gotalignfaults == 0, "got %d alignment faults, expected 0\n", gotalignfaults); +#endif + + pNtCurrentTeb()->Tib.ExceptionList = exc_frame.frame.Prev; +} + #endif /* __i386__ */ START_TEST(exception) @@ -288,5 +431,7 @@ START_TEST(exception) #ifdef __i386__ test_prot_fault(); test_debug_regs(); + test_single_step(); + test_align_faults(); #endif } -- 1.2.4

Oct. 2, 2006

Re: no backtrace with wine crashes

by James Hawkins

On 10/2/06, Eric Pouech <eric.pouech(a)wanadoo.fr> wrote: > James Hawkins wrote: > > > Auto = '1' > > Debugger = 'winedbg --auto %ld %ld' > > > does the attached patch help ? > > > [WineDbg]: fixed the auto mode > > From: Eric Pouech <eric.pouech(a)wanadoo.fr> > I had to apply it by hand, but once I did it worked. Thanks Eric. -- James Hawkins

Oct. 2, 2006

Re: no backtrace with wine crashes

by Eric Pouech

James Hawkins wrote: > Auto = '1' > Debugger = 'winedbg --auto %ld %ld' > does the attached patch help ? [WineDbg]: fixed the auto mode From: Eric Pouech <eric.pouech(a)wanadoo.fr> --- programs/winedbg/info.c | 2 +- programs/winedbg/tgt_active.c | 6 +++++- 2 files changed, 6 insertions(+), 2 deletions(-) diff --git a/programs/winedbg/info.c b/programs/winedbg/info.c index d2042bf..c5b7b1e 100644 --- a/programs/winedbg/info.c +++ b/programs/winedbg/info.c @@ -204,7 +204,7 @@ void info_win32_module(DWORD base) int i, j, num_printed = 0; DWORD opt; - if (!dbg_curr_process || !dbg_curr_thread) + if (!dbg_curr_process) { dbg_printf("Cannot get info on module while no process is loaded\n"); return; diff --git a/programs/winedbg/tgt_active.c b/programs/winedbg/tgt_active.c index af619a0..48f1ec9 100644 --- a/programs/winedbg/tgt_active.c +++ b/programs/winedbg/tgt_active.c @@ -867,7 +867,8 @@ enum dbg_start dbg_active_auto(int argc, argc--; argv++; ds = dbg_active_attach(argc, argv); if (ds != start_ok) return ds; - hFile = parser_generate_command_file("echo Modules:", "info share", + hFile = parser_generate_command_file("backtrace", + "echo Modules:", "info share", "echo Threads:", "info threads", NULL); } @@ -923,6 +924,9 @@ enum dbg_start dbg_active_auto(int argc, else return start_error_parse; if (hFile == INVALID_HANDLE_VALUE) return start_error_parse; + if (dbg_curr_process->active_debuggee) + dbg_active_wait_for_first_exception(); + dbg_interactiveP = TRUE; parser_handle(hFile);

Oct. 2, 2006

Re: no backtrace with wine crashes

by James Hawkins

On 10/2/06, Eric Pouech <eric.pouech(a)wanadoo.fr> wrote: > James Hawkins wrote: > > > Hey guys, > > > > Since very recently, any time we get a crash in Wine, winedbg does not > > show a backtrace. This makes it significantly more difficult to debug > > the crash. Is anyone else experiencing this? I'll run a regression > > test as soon as I can. > > > what's the format of your AeDebug registry key ? > it seems it tries to start a program while it's supposed to attach to a > running program > A+ > Auto = '1' Debugger = 'winedbg --auto %ld %ld' -- James Hawkins

Oct. 2, 2006

Re: no backtrace with wine crashes

by Eric Pouech

James Hawkins wrote: > Hey guys, > > Since very recently, any time we get a crash in Wine, winedbg does not > show a backtrace. This makes it significantly more difficult to debug > the crash. Is anyone else experiencing this? I'll run a regression > test as soon as I can. what's the format of your AeDebug registry key ? it seems it tries to start a program while it's supposed to attach to a running program A+

Oct. 2, 2006

Implementation question for CryptSIPLoad

by Paul Vriens

Hi, In CryptSIPLoad I have to load the some internal crypt32 function pointers into a structure. Will the following work ? : pSipDispatch->pfGet = &CryptSIPGetSignedDataMsg; pSipDispatch->pfPut = &CryptSIPPutSignedDataMsg; pSipDispatch->pfCreate = &CryptSIPCreateIndirectData; pSipDispatch->pfVerify = &CryptSIPVerifyIndirectData; pSipDispatch->pfRemove = &CryptSIPRemoveSignedDataMsg; What happens if I have another dll loaded that has the same functions (wintrust, msisip ....)? Do I need to do an explicit LoadLibrary or GetModuleHandle? The tests I've already created (in current GIT) succeed with the above 'solution' but I just want to make sure there aren't any side-effects. Cheers, Paul.

Oct. 2, 2006

Re: no backtrace with wine crashes

by James Hawkins

On 10/2/06, James Hawkins <truiken(a)gmail.com> wrote: > Hey guys, > > Since very recently, any time we get a crash in Wine, winedbg does not > show a backtrace. This makes it significantly more difficult to debug > the crash. Is anyone else experiencing this? I'll run a regression > test as soon as I can. > Update: running apps that crash in winedbg gives the proper backtrace. -- James Hawkins

Oct. 2, 2006

Re: compiling wine for amd64 on ubuntu dapper 6.06

by Evil Jay

It's very useful - CounterStrike source is actually playable with it. I'm sure other games are working as well. -J Ge van Geldorp wrote: >> From: "Gerald Britton" <gerald.britton(a)gmail.com> >> >> 1. /configure can't find opengl, and produces these messages: >> >> configure: WARNING: Wine will be build without OpenGL or >> Direct3D support >> configure: WARNING: because something is wrong with the OpenGL setup: >> configure: WARNING: No OpenGL library found on this system. >> > > I found the easiest way around this problem is to run configure using > --without-opengl. I doubt OpenGL is very useful at the moment for 64-bit > Wine anyway. > > Ge van Geldorp. > > > > > >

Oct. 2, 2006

no backtrace with wine crashes

by James Hawkins

Hey guys, Since very recently, any time we get a crash in Wine, winedbg does not show a backtrace. This makes it significantly more difficult to debug the crash. Is anyone else experiencing this? I'll run a regression test as soon as I can. wine: Unhandled page fault on read access to 0x00000000 at address 0x7ebfeb70 (thread 0011), starting debugger... Modules: Cannot get info on module while no process is loaded Threads: process tid prio (all id:s are in hex) 00000013 00000021 0 00000020 0 0000001e 0 0000001b 0 00000019 0 00000018 0 00000017 0 00000016 0 00000014 0 00000010 (D) (null) 00000023 0 00000022 0 00000012 0 00000011 0 0000000a 0000000c 0 0000000b 0 -- James Hawkins

Oct. 2, 2006