July 2004 - Wine-Devel - list.winehq.org

Re: Question about mmap issues on FreeBSD/NetBSD

by Alexandre Julliard

Yorick Hardy <yha(a)na.rau.ac.za> writes: > Also in libs/wine/mmap.c line 119 the result of mincore is never tested, > should it not be > > if ( (mincore( (caddr_t)addr + i, pagesize, &vec ) != -1) && (vec&1) ) No, the goal is to check if something is mapped there, we don't actually care whether it's in core or swapped out. -- Alexandre Julliard julliard(a)winehq.org

July 30, 2004

Re: RPCRT4_NdrClientCall2 stub (fixme)?

by Robert Shearman

Vincent Béron wrote: >Le ven 30/07/2004 à 11:38, Robert Shearman a écrit : >[snip] > > >>>>It wasn't Alexandre's fault. I didn't ask for it to be merged. One of >>>>the main problems is that because of calling convention differences for >>>>passing unions/structures we can't be binary compatible with Windows at >>>>the same time as being source compatible. >>>> >>>> >>>> >>>> >>>You can take a glance at our msvcrt div/ldiv implementation which is >>>both source and binary compatible. >>>binary: dlls/msvcrt/math.c >>>source: include/msvcrt/stdlib.h (which then calls dlls/msvcrt/math.c) >>> >>> >>> >>> >>NdrClientCall2 is a varargs function, so I don't think that approach >>would work. >> >> > >Sorry, misread the "passing unions/structures" part as "returning >unions/structures". > > No, I did mean "returing unions/structures" (the parameters don't matter as all the complicated ones must be passed by reference). Rob

July 30, 2004

Re: for whoever maintinas winrash

by Chris Morgan

Thanks for the bug report. On Friday 30 July 2004 12:46 am, Ivan Leo Puoti wrote: > We've had the following complaints on wine-users. > > > I installed this on a windows 98se box. I can't say I like having the DOS > > window open all the time (it's ugly). Can this not be written as a > > windows programme, and sit in the system tray? > > I'm not really sure what the user means. I don't have anything other than an xp machine to test on and services are supported there. Anyone know what the issue is? Maybe adding some code to the non-service path that was added to support win9X machines would let us hide the window? > > Also, during the installation process, after I entered my email address, > > the installer removed what it called 'invalid characters' - i.e., > > underscores. All my main email addresses contain underscores, so I was > > forced to enter an old one that is not very reliable. We could support underscores but Feri tells me that they aren't supported by winetest so I strip them out both during the install and when the email address is being used. Feri, anyway we can add support for underscores in winetest and remove this requirement? Chris

July 30, 2004

Re: possible bug in wcmd

by Ivan Leo Puoti

It looks like a bug to me. [ivan(a)localhost wcmd]$ ./wcmd WCMD Version 0.17 Z:\home\ivan\Development\Wine\CVS\wine\wine\programs\wcmd>help for wine: Unhandled exception (thread 0009), starting debugger... fixme:console:SetConsoleCtrlHandler (0x405ec9e0,1) - no error checking or testing yet WineDbg starting on pid 0x8 Unhandled exception: page fault on read access to 0x20656854 in 32-bit code (0x400d2473). In 32 bit mode. 0x400d2473 strlen in libc.so.6: movl 0x0(%eax),%ecx Wine-dbg>bt Backtrace: =>1 0x400d2473 strlen+0x33 in libc.so.6 (0x406eed1c) 2 0x400ba8cb _IO_vsprintf+0x8b in libc.so.6 (0x406eede4) 3 0x405c9eeb WCMD_output+0x2b(format=0x406ef20c) [wcmdmain.c:741] in wcmd (0x406ef200) 4 0x405c7020 WCMD_give_help+0x80(command=0x4039b8a5) [builtins.c:352] in wcmd (0x406efa18) 5 0x405c9563 .L153+0x9 in wcmd (0x406efa58) 6 0x405c91d4 main+0x4f4(argc=0x1, argv=0xbffff49c) [wcmdmain.c:244] in wcmd (0x406efe9c) 7 0x405c6169 __wine_exe_main+0x169 in wcmd (0x406eff2c) 8 0x404e6a88 start_process+0xc8(arg=0x0) [process.c:995] in kernel32 (0x406efff4) 9 0x4001cd9d wine_switch_to_stack+0x11 in libwine.so.1 (0x00000000) Wine-dbg>quit WineDbg terminated on pid 0x8 [ivan(a)localhost wcmd]$

July 30, 2004

Re: RPCRT4_NdrClientCall2 stub (fixme)?

by Mike Hearn

On Fri, 30 Jul 2004 17:48:08 +0200, Jeroen Janssen wrote: > Hmmm, so it can (unfortunately) happen that a patch is 'forgotten' Yes that does happen sometimes. It's up to the author to resubmit it. Alternatively sometimes a patch is correct but can't be applied for other reasons, like my lockwindowupdate patch or systray patch. > Oh, well.. I actually ment to compare 'one person with CVS write > permissions' to 'multiple persons with CVS write permissions'. > > In the past, (open source) projects I worked on had multiple persons > with CVS write access to the repository. That is more typical yes. > Related to the kernel, some people keep their 'personal' tree (like -mm, > -ac, etc) that can be used to structurally test certain patches, etc > and then subsequently patches can be gradially merged from these trees > to the 'official main tree' (after some testing). > > Is this something that also would work for wine? --- Jeroen Maybe, that is what arch was all about, but it's not really mature enough to work well in Wine yet I think. External patchsets might happen at some point, I don't know. I think most people are happy with the current system.

July 30, 2004

possible bug in wcmd

by Henning Gerhardt

Hi, I discovered a possible bug in wcmd. If you run "help for" inside wcmd, wcmd crashed and the winedebugger is called. I used to confirm this bug on my machine the latest tarball and a CVS-snapshot from today. If anybody can confirm this bug I will be happy :) The following lines are only correct if this bug is true: The problem is that a single % letter inside the resource strings. Look for the for-help text in En.rc file and you find the sentence "The requirement to double the % sign when using FOR in a batch file does") get parsed and this single % caused the crash. I found the following solutions and I don't know what is the right solution: 1. replace the single % with a double %% in all resource files 2. replace in the function WCMD_give_help (builtins.c:330) WCMD_output with WCMD_output_asis (WCMD_output (wcmdmain.c:733) called WCMD_output_asis but make some stuff for possible dynamic parameters) 3. patch the function WCMD_output (wcmdmain.c:733) with the attached patch (Replace a single % with a double % and all later called system function are be happy) I prefer personal the second or third solution. Regards, Henning

July 30, 2004

Re: RPCRT4_NdrClientCall2 stub (fixme)?

by Vincent Béron

Le ven 30/07/2004 à 11:38, Robert Shearman a écrit : [snip] > >>It wasn't Alexandre's fault. I didn't ask for it to be merged. One of > >>the main problems is that because of calling convention differences for > >>passing unions/structures we can't be binary compatible with Windows at > >>the same time as being source compatible. > >> > >> > > > >You can take a glance at our msvcrt div/ldiv implementation which is > >both source and binary compatible. > >binary: dlls/msvcrt/math.c > >source: include/msvcrt/stdlib.h (which then calls dlls/msvcrt/math.c) > > > > > > NdrClientCall2 is a varargs function, so I don't think that approach > would work. Sorry, misread the "passing unions/structures" part as "returning unions/structures". Still, if you're able to have both (a source compatible and a binary compatible) independantly, there's probably a clever way to have the source one call the binary one. Vincent

July 30, 2004

Re: RPCRT4_NdrClientCall2 stub (fixme)?

by Jeroen Janssen

Mike Hearn wrote: >>Is there an official way to track that a patch submitted to wine-patches >>has been comitted? Also can I somehow find out which (interesting) >>patches have not been comitted yet? > > Unfortunately there is no way to track this in an automated fashion. It > has been talked about but nobody created a system yet. We just watch > wine-cvs and see when it's been checked in. Hmmm, so it can (unfortunately) happen that a patch is 'forgotten' >>Also, I noticed there is currently only one person with CVS write >>permissions for Wine? Has this always been the case? > > Yes. Before we had CVS it was like the kernel, people just mailed > Alexandre patches and didn't know if they'd been merged until the next > release (as far as I know, but that was long before my time, back then I > thought Win98 was the bees knees ;) Oh, well.. I actually ment to compare 'one person with CVS write permissions' to 'multiple persons with CVS write permissions'. In the past, (open source) projects I worked on had multiple persons with CVS write access to the repository. Related to the kernel, some people keep their 'personal' tree (like -mm, -ac, etc) that can be used to structurally test certain patches, etc and then subsequently patches can be gradially merged from these trees to the 'official main tree' (after some testing). Is this something that also would work for wine? --- Jeroen

July 30, 2004

Re: RPCRT4_NdrClientCall2 stub (fixme)?

by Robert Shearman

Vincent Béron wrote: >Le ven 30/07/2004 à 07:39, Robert Shearman a écrit : > > >>Mike Hearn wrote: >> >> >> >>>On Thu, 29 Jul 2004 19:46:18 +0100, Robert Shearman wrote: >>> >>> >>> >>> >>>>Yes, I have an implementation of NdrClientCall2 in my local tree, but >>>>I'm a bit busy with other COM things at the moment. I'll send an updated >>>>patch to the list either tonight or sometime over the weekend. >>>> >>>> >>>> >>>> >>>Alexandre, why was this not merged when it was originally sent? >>> >>> >>> >>> >>It wasn't Alexandre's fault. I didn't ask for it to be merged. One of >>the main problems is that because of calling convention differences for >>passing unions/structures we can't be binary compatible with Windows at >>the same time as being source compatible. >> >> > >You can take a glance at our msvcrt div/ldiv implementation which is >both source and binary compatible. >binary: dlls/msvcrt/math.c >source: include/msvcrt/stdlib.h (which then calls dlls/msvcrt/math.c) > > NdrClientCall2 is a varargs function, so I don't think that approach would work. Rob

July 30, 2004

list view completely mangled in WINE

by Tobias Burnus

Hello, in Crystal Impact's Diamond the list view is completely mangled and only barely recognizable as list view. Result in WINE: http://www.physik.fu-berlin.de/~tburnus/wine/diamond_list_wine.png Result in Windows: http://www.physik.fu-berlin.de/~tburnus/wine/diamond_list.png I'd be really pleased if someone could look at the problem. The demo version of the program is free and available from http://www.crystalimpact.com/download/diamond/demo/dm2demo.exe (In addition to this list-view problem the application does off-screen rendering which is not supported by WINE; one needs to disable rendering in order use the program. See WINE bug http://bugs.winehq.com/show_bug.cgi?id=2320 ) Tobias

July 30, 2004