September 2003 - Wine-Devel

Re: [winecfg 8] Complete drive edit dialog, warn on lack of C drive, implement drive delete, code cleanups
by Dimitrie O. Paun Sept. 12, 2003

Sept. 12, 2003

On 11 Sep 2003, Mike Hearn wrote: > ChangeLog: > - All settings in the drive edit dialog are now instant apply This is a cool feature, but we should keep it optional. Our primary audience are Windows users, and they are not used to that. Heck, even I am not used to it, I am used to be able to play around with stuff in dialogs, and than press CANCEL. In fact, KDE is not instant apply, so I would argue that even most Linux users are used to that. I'd argue that we need to support instant apply to integrate with GNOME, but to default to the classic OK/CANCEL for now, until we can figure how and when to enable one or the other behaviour. -- Dimi.

5 8

Kernel pipe test segfaults?
by Mike Hearn Sept. 12, 2003

Sept. 12, 2003

Maybe this is a known issue, I haven't really followed the test case work up until now, but does the kernel32 pipe test segfault for anybody else? thanks -mike

2 1

Re: WIne help - get rid of W->A calls
by Eric Pouech Sept. 12, 2003

Sept. 12, 2003

hatky wrote: > My first patch, so a deep check is welcome > > Change log: > - Remove W->A call from WineHelpA/W > - This patch is dedicated to dimi that guided me I'm afraid the patch is wrong. the current implentation of WinHelp, which sends data from any app to the program winhelp, requires that the data block passed between those two apps encodes strings in ASCII, not unicode. Your patch modifies the sender so that it encodes in Unicode, but the receiver is left untouched. I'd suggest anyway that we remain as close as possible as Windows do. - we don't completly handle this protocol as Windows do (Windows send a WM_HELP - registered - message, we only use a WM_COPYDATA), but the format of the message (parameter block) is quite the same (we just add some extra fields to ensure that we don't mess with a spurious WM_COPYDATA message in winhelp). - as Windows encode strings in this message block in ASCII, I'd suggest keeping the encoding of strings inside the message as ASCII. A+ -- Eric Pouech

3 2

configure new wine
by eric Sept. 12, 2003

Sept. 12, 2003

Dear cvs wine users: By reading the doc's install page, I success uninstall cvs 's wine reinstall 8/13' wine, then test again, it indeed repeat the ikernel problem, but it at least configure correct by default. So I copy /root/.wine/config to /root/oldwineconfig. Then uninstall it again as doc indicate, then install cvs again, after that I cp oldwineconfig /etc/wine/config and /root/.wine/config then run my app, it still show segmentation fault please help, thanks in advance Sincerely Eric

1 0

Can not uninstall CVS wine
by eric Sept. 12, 2003

Sept. 12, 2003

Dear Wine users: After I install CVS wine, and find it is not work, it type rpm -e wine reinstall my old wine, then if I did not do it by force, it showed error ------------------ [root(a)localhost x86]# rpm -vhi /root/wine-2* Preparing... ########################################### [100%] file /usr/bin/widl from install of wine-20030813-1rh9winehq conflicts with file from package wine-cvs-unstripped-20030911-1 file /usr/bin/wine from install of wine-20030813-1rh9winehq conflicts with file from package wine-cvs-unstripped-20030911-1 ------------------ if I use force to install old package, it still can not run my app, segmentation fault, and I check rpm -q wine, it reponse wine package in not installed highly appreciate any help to remove that CVS's wine throughly and thanks in advance Sincerely Eric

1 0

openbsd supports MAP_FIXED too
by Ted Unangst Sept. 12, 2003

Sept. 12, 2003

sorry no patch. i'm not sure what the status of wine on openbsd is, although i came across a notice that wine was capable of using MAP_FIXED. i committed what amounted to todd's (netbsd) patch to openbsd shortly after it went into netbsd. don't know if this helps / who cares, but fyi. patch below could probably be || defined (__OpenBSD__) as well. http://www.winehq.org/hypermail/wine-patches/2003/08/0278.html -- someone's writing down your mistakes someone's documenting your downfall

1 0

scheduler related complete breakage on FreeBSD
by Gerald Pfeifer Sept. 12, 2003

Sept. 12, 2003

On FreeBSD 4.8 (using -lc_r instead of -lc to get pthreads) Wine crashes upon startup: /sw/test/wine% wine FIXME:pthread_cond_init Segmentation fault (core dumped) % gdb53 miscemu/wine wine.core GNU gdb 5.3 (FreeBSD) Reading symbols from /sw/test/wine/dlls/libntdll.dll.so...done. Loaded symbols for /sw/wine/dlls/libntdll.dll.so Reading symbols from /sw/test/wine/libs/libwine.so.1...done. Loaded symbols for /sw/test/wine/libs/libwine.so.1 Reading symbols from /sw/test/wine/libs/libwine_unicode.so.1...done. Loaded symbols for /sw/test/wine/libs/libwine_unicode.so.1 Reading symbols from /usr/lib/libc.so.4...done. Loaded symbols for /usr/lib/libc.so.4 Reading symbols from /usr/lib/libm.so.2...done. Loaded symbols for /usr/lib/libm.so.2 Reading symbols from /usr/lib/libc_r.so.4...done. Loaded symbols for /usr/lib/libc_r.so.4 Reading symbols from /sw/test/wine/dlls/kernel32.dll.so...done. Loaded symbols for /sw/test/wine/dlls/kernel32.dll.so Reading symbols from /usr/libexec/ld-elf.so.1...done. Loaded symbols for /usr/libexec/ld-elf.so.1 (gdb) bt #0 0x5c251aac in .cerror () from /usr/lib/libc.so.4 #1 0x5c2feca6 in flock () from /usr/lib/libc_r.so.4 #2 0x5c231919 in _yp_dobind () from /usr/lib/libc.so.4 #3 0x5c231fa8 in yp_match () from /usr/lib/libc.so.4 #4 0x5c215c12 in endpwent () from /usr/lib/libc.so.4 #5 0x5c2153b8 in getpwuid () from /usr/lib/libc.so.4 #6 0x5c0d06d1 in init_paths () at config.c:112 #7 0x5c0d0a4e in wine_get_server_dir () at config.c:191 #8 0x5c07e5ae in CLIENT_InitServer () at ../../scheduler/client.c:641 #9 0x5c07f4f4 in process_init (argv=0xbfbff690) at ../../scheduler/process.c:306 #10 0x5c07fa19 in __wine_process_init (argc=1, argv=0xbfbff690) at ../../scheduler/process.c:480 #11 0x5c0d21c4 in wine_init (argc=1, argv=0xbfbff690, error=0xbfbff240 "", error_size=1024) at loader.c:428 #12 0x3c000621 in main (argc=1, argv=0xbfbff690) at main.c:33 #13 0x3c000566 in _start () Any help will be appreciated. Gerald

1 1

FIXME:pthread_cond_init
by Gerald Pfeifer Sept. 12, 2003

Sept. 12, 2003

When building Wine on FreeBSD with libc_r (r for reentrant) instead of libc, I now trigger the following FIXME in scheduler/pthread.c: /***** CONDITIONS *****/ /* not implemented right now */ int __pthread_cond_init(pthread_cond_t *cond, const pthread_condattr_t *cond_attr) { P_OUTPUT("FIXME:pthread_cond_init\n"); return 0; } What's the proper way to address this? (I'm afraid, it's beyond what I'll be able to implement, though.) Gerald

1 1

BuildCommDCB conformance test
by Kevin Groeneveld Sept. 12, 2003

Sept. 12, 2003

Here is my first attempt at a conformance test for the BuildCommDCB functions. It does not cover very many cases yet, but I thought I would post what I have so far in case I am going about this the wrong way. I compiled this with MSVC and it ran with no errors on both Windows 95 and Windows 2000. It also compiles with libwine in Linux, but for this to work you need to add a missing member to the DCB structure in winbase.h. If I run this test with the existing implementation of BuildCommDCB it page faults, probably because it is writing to a constant string. If my previous patch to BuildCommDCB is applied, then the test runs ok. Kevin diff -urN wine-20030813/dlls/kernel/tests/Makefile.in wine-20030813.patched/dlls/kernel/tests/Makefile.in --- wine-20030813/dlls/kernel/tests/Makefile.in 2003-08-23 23:12:31.000000000 -0400 +++ wine-20030813.patched/dlls/kernel/tests/Makefile.in 2003-09-10 23:31:59.000000000 -0400 @@ -8,6 +8,7 @@ CTESTS = \ alloc.c \ atom.c \ + build_dcb.c \ codepage.c \ console.c \ directory.c \ diff -urN wine-20030813/dlls/kernel/tests/build_dcb.c wine-20030813.patched/dlls/kernel/tests/build_dcb.c --- wine-20030813/dlls/kernel/tests/build_dcb.c 1969-12-31 19:00:00.000000000 -0500 +++ wine-20030813.patched/dlls/kernel/tests/build_dcb.c 2003-09-11 01:00:19.000000000 -0400 @@ -0,0 +1,171 @@ +/* Unit test suite for BuildCommDCBA/W & BuildCommDCBAndTimeoutsA/W + * + * Copyright 2003 Kevin Groeneveld + * + * This library is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; either + * version 2.1 of the License, or (at your option) any later version. + * + * This library is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this library; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#include "wine/test.h" +#include "winbase.h" + +typedef struct +{ + char *string; + BOOL result; + DCB dcb1, dcb2; + COMMTIMEOUTS timeouts1, timeouts2; +} TEST; + +TEST test[] = +{ + { + "baud=9600", + TRUE, + { 0x00000000, 9600, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0x00000, 0x0000, 0x0000, 0x0000, 0x00, 0x00, 0x00, (char)0x00, (char)0x00, (char)0x00, (char)0x00, (char)0x00, 0x0000 }, + { 0xffffffff, 9600, 1, 1, 1, 1, 3, 1, 1, 1, 1, 1, 1, 3, 1, 0x1ffff, 0xffff, 0xffff, 0xffff, 0xff, 0xff, 0x00, (char)0xff, (char)0xff, (char)0xff, (char)0xff, (char)0xff, 0xffff }, + { 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000 }, + { 0xffffffff, 0xffffffff, 0xffffffff, 0xffffffff, 0xffffffff } + }, + { + " baud=9600", + FALSE, + { 0x00000000, 0x00000000, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0x00000, 0x0000, 0x0000, 0x0000, 0x00, 0x00, 0x00, (char)0x00, (char)0x00, (char)0x00, (char)0x00, (char)0x00, 0x0000 }, + { 0xffffffff, 0xffffffff, 1, 1, 1, 1, 3, 1, 1, 1, 1, 1, 1, 3, 1, 0x1ffff, 0xffff, 0xffff, 0xffff, 0xff, 0xff, 0xff, (char)0xff, (char)0xff, (char)0xff, (char)0xff, (char)0xff, 0xffff }, + { 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000 }, + { 0xffffffff, 0xffffffff, 0xffffffff, 0xffffffff, 0xffffffff } + }, + { + "to=on", + TRUE, + { 0x00000000, 0x00000000, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0x00000, 0x0000, 0x0000, 0x0000, 0x00, 0x00, 0x00, (char)0x00, (char)0x00, (char)0x00, (char)0x00, (char)0x00, 0x0000 }, + { 0xffffffff, 0xffffffff, 1, 1, 1, 1, 3, 1, 1, 1, 1, 1, 1, 3, 1, 0x1ffff, 0xffff, 0xffff, 0xffff, 0xff, 0xff, 0x00, (char)0xff, (char)0xff, (char)0xff, (char)0xff, (char)0xff, 0xffff }, + { 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x0000EA60 }, + { 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x0000EA60 } + }, +}; + +#define TEST_COUNT (sizeof(test) / sizeof(TEST)) + +void check_result(char *function, char *string, int initial_value, BOOL result1, BOOL result2) +{ + DWORD LastError = GetLastError(); + DWORD CorrectError = (result2 ? 0xdeadbeef : ERROR_INVALID_PARAMETER); + + ok(LastError == CorrectError, "%s(\"%s\"), 0x%02x: GetLastError() returned 0x%08lx, should be 0x%08lx", function, string, initial_value, LastError, CorrectError); + ok(result1 == result2, "%s(\"%s\"), 0x%02x: return value should be %s", function, string, initial_value, result2 ? "TRUE" : "FALSE"); +} + +#define check_dcb_member(a,b) ok(pdcb1->a == pdcb2->a, "%s(\"%s\"), 0x%02x: "#a" is "b", should be "b, function, string, initial_value, pdcb1->a, pdcb2->a); + +void check_dcb(char *function, char *string, int initial_value, DCB *pdcb1, DCB *pdcb2) +{ + /* DCBlength is a special case since Win 9x sets it but NT does not. + We will accept either as correct. */ + ok(pdcb1->DCBlength == pdcb2->DCBlength || pdcb1->DCBlength == sizeof(DCB), "%s(\"%s\"), 0x%02x: DCBlength is %lu, should be %lu or %lu", function, string, initial_value, pdcb1->DCBlength, pdcb2->DCBlength, (DWORD)sizeof(DCB)); + + /* Check all other DCB members */ + check_dcb_member(BaudRate, "%lu"); + check_dcb_member(fBinary, "%u"); + check_dcb_member(fParity, "%u"); + check_dcb_member(fOutxCtsFlow, "%u"); + check_dcb_member(fOutxDsrFlow, "%u"); + check_dcb_member(fParity, "%u"); + check_dcb_member(fOutxCtsFlow, "%u"); + check_dcb_member(fOutxDsrFlow, "%u"); + check_dcb_member(fDtrControl, "%u"); + check_dcb_member(fDsrSensitivity, "%u"); + check_dcb_member(fTXContinueOnXoff, "%u"); + check_dcb_member(fOutX, "%u"); + check_dcb_member(fInX, "%u"); + check_dcb_member(fErrorChar, "%u"); + check_dcb_member(fNull, "%u"); + check_dcb_member(fRtsControl, "%u"); + check_dcb_member(fAbortOnError, "%u"); + check_dcb_member(fDummy2, "%u"); + check_dcb_member(wReserved, "%u"); + check_dcb_member(XonLim, "%u"); + check_dcb_member(XoffLim, "%u"); + check_dcb_member(ByteSize, "%u"); + check_dcb_member(Parity, "%u"); + check_dcb_member(StopBits, "%u"); + check_dcb_member(XonChar, "%d"); + check_dcb_member(XoffChar, "%d"); + check_dcb_member(ErrorChar, "%d"); + check_dcb_member(EofChar, "%d"); + check_dcb_member(EvtChar, "%d"); + check_dcb_member(wReserved1, "%u"); +} + +#define check_timeouts_member(a) ok(ptimeouts1->a == ptimeouts2->a, "%s(\"%s\"), 0x%02x: "#a" is %lu, should be %lu", function, string, initial_value, ptimeouts1->a, ptimeouts2->a); + +void check_timeouts(char *function, char *string, int initial_value, COMMTIMEOUTS *ptimeouts1, COMMTIMEOUTS *ptimeouts2) +{ + check_timeouts_member(ReadIntervalTimeout); + check_timeouts_member(ReadTotalTimeoutMultiplier); + check_timeouts_member(ReadTotalTimeoutConstant); + check_timeouts_member(WriteTotalTimeoutMultiplier); + check_timeouts_member(WriteTotalTimeoutConstant); +} + +void test_BuildCommDCBA(char* string, BOOL expected_result, DCB *pexpected_dcb, int initial_value) +{ + BOOL result; + DCB dcb; + + /* set initial conditions */ + memset(&dcb, initial_value, sizeof(DCB)); + SetLastError(0xdeadbeef); + + result = BuildCommDCBA(string, &dcb); + + /* check results */ + check_result("BuildCommDCBA", string, initial_value, result, expected_result); + check_dcb("BuildCommDCBA", string, initial_value, &dcb, pexpected_dcb); +} + +void test_BuildCommDCBAndTimeoutsA(char* string, BOOL expected_result, DCB *pexpected_dcb, COMMTIMEOUTS *pexpected_timeouts, int initial_value) +{ + BOOL result; + DCB dcb; + COMMTIMEOUTS timeouts; + + /* set initial conditions */ + memset(&dcb, initial_value, sizeof(DCB)); + memset(&timeouts, initial_value, sizeof(COMMTIMEOUTS)); + SetLastError(0xdeadbeef); + + result = BuildCommDCBAndTimeoutsA(string, &dcb, &timeouts); + + /* check results */ + check_result("BuildCommDCBAndTimeoutsA", string, initial_value, result, expected_result); + check_dcb("BuildCommDCBAndTimeoutsA", string, initial_value, &dcb, pexpected_dcb); + check_timeouts("BuildCommDCBAndTimeoutsA", string, initial_value, &timeouts, pexpected_timeouts); +} + +void do_test(TEST *ptest) +{ + test_BuildCommDCBA(ptest->string, ptest->result, &ptest->dcb1, 0x00); + test_BuildCommDCBA(ptest->string, ptest->result, &ptest->dcb2, 0xFF); + test_BuildCommDCBAndTimeoutsA(ptest->string, ptest->result, &ptest->dcb1, &ptest->timeouts1, 0x00); + test_BuildCommDCBAndTimeoutsA(ptest->string, ptest->result, &ptest->dcb2, &ptest->timeouts2, 0xFF); +} + +START_TEST(build_dcb) +{ + int i; + + for(i = 0; i < TEST_COUNT; i++) + do_test(&test[i]); +} diff -urN wine-20030813/include/winbase.h wine-20030813.patched/include/winbase.h --- wine-20030813/include/winbase.h 2003-08-23 23:12:51.000000000 -0400 +++ wine-20030813.patched/include/winbase.h 2003-09-08 20:19:39.000000000 -0400 @@ -976,6 +976,7 @@ char ErrorChar; char EofChar; char EvtChar; + WORD wReserved1; } DCB, *LPDCB; typedef struct tagCOMMCONFIG {

4 4

Re: missing ordinals in avifil32
by Alexandre Julliard Sept. 12, 2003

Sept. 12, 2003

Dave Miller <compsol(a)ptd.net> writes: > The first 59 ordinals seem to match in all windows versions. Add > ordinals to avifil32.spec to match windows. Unless ordinals are actually needed, it's better not to add them. In that case the functions are sorted so it's very likely that Windows is using automatic allocation too. -- Alexandre Julliard julliard(a)winehq.com

3 3