weiden(a)cvs.reactos.com wrote:
>CVSROOT: /CVS/ReactOS
>Module name: reactos
>Repository: reactos/subsys/win32k/stubs/
>Changes by: weiden(a)mok.osexperts.com 04/12/11 17:40:44
>
>Modified files:
> ./: config
> reactos/include/napi/: win32.h
> reactos/include/win32k/: bitmaps.h brush.h dc.h gdiobj.h kapi.h
> ntuser.h pen.h region.h text.h
> reactos/lib/user32/include/: user32.h window.h
> reactos/lib/user32/misc/: dllmain.c stubs.c
> reactos/lib/user32/windows/: defwnd.c
> reactos/subsys/win32k/eng/: bitblt.c clip.c mouse.c objects.h
> surface.c xlate.c
> reactos/subsys/win32k/include/: cursoricon.h dce.h inteng.h
> intgdi.h mouse.h object.h
> palette.h tags.h
> reactos/subsys/win32k/main/: dllmain.c
> reactos/subsys/win32k/: makefile w32k.h
> reactos/subsys/win32k/misc/: object.c
> reactos/subsys/win32k/ntddraw/: ddraw.c
> reactos/subsys/win32k/ntuser/: desktop.c input.c misc.c stubs.c
> windc.c window.c winpos.c
> winsta.c
> reactos/subsys/win32k/objects/: bitmaps.c brush.c cliprgn.c
> color.c dc.c dib.c fillshap.c
> gdiobj.c line.c palette.c pen.c
> print.c region.c text.c
> reactos/subsys/win32k/stubs/: stubs.c xpstubs.c
>Added files:
> reactos/subsys/win32k/include/: intddraw.h
> reactos/subsys/win32k/ntuser/: cursoricon.c
> reactos/subsys/win32k/objects/: stockobj.c
>Removed files:
> reactos/subsys/win32k/eng/: clip.h handle.c handle.h
> reactos/subsys/win32k/objects/: cursoricon.c
>
>Log message:
> 1. remove mouse cursor hacks from eng/mouse
> 2. implemented more win2k-like gdi handle table (handle entry lookup is not yet optimized/cached!)
> 3. moved cursoricon.c from objects to ntuser as it's part of the user api and not of the gdi
> 4. get rid of the eng handle table
> 5. minor fixes
>
>
I'm not sure if it's exactly this commit, but Miranda IM now crashes
win32k badly. Also the mouse cursor hacks had a reason... if you remove
them, please fix the problem different way. Now there is no cursor set
during the startup and so when Explorer tries to revert the hourglass
cursor to a normal one, the call just fails.
Regards,
Filip
While trying to install a current CVS (both freeldr and reactos)
bootcd on real hardware, I was greeted with a DIVIDE BY ZERO
exception. After asking some people on IRC and poking around in
Bugzilla, It turned out I was not getting all the output I needed, due
to what seems to be doublespaced output..
Screenshot: http://waxdragon.homeip.net/~ford/freeldr_crash.png
After poking around a bit, I rolled i386trap.S back to rev 1.5 and
then I got the second half of the output. I stared at i386trap.S for
a while, and cannot figure out where the problem is.
FreeLoader gurus, can you help me, please? ;0)
WD
--
The cheese stands alone.
Hi all!
weiden(a)cvs.reactos.com wrote:
> CVSROOT: /CVS/ReactOS
> Module name: reactos
> Repository: reactos/lib/advapi32/sec/
> Changes by: weiden(a)mok.osexperts.com 04/12/10 16:21:34
>
> Modified files:
> reactos/lib/advapi32/: advapi32.def advapi32.h makefile
> reactos/lib/advapi32/misc/: sysfunc.c
> reactos/lib/advapi32/token/: token.c
> Added files:
> reactos/lib/advapi32/crypt/: .cvsignore crypt.c crypt.h
> crypt_des.c crypt_lmhash.c
> crypt_md4.c crypt_md5.c crypt_sha.c
> reactos/lib/advapi32/sec/: trustee.c
>
> Log message:
> port crypt and trustee support from Wine. Patch by Steven Edwards.
I'm getting this,
[CC] crypt/crypt.c
In file included from crypt/crypt.c:28:
../../include/wine/port.h:248: error: parse error before "pread"
../../include/wine/port.h:248: warning: type defaults to `int' in declaration of `pread'
../../include/wine/port.h:248: warning: data definition has no type or storage class
../../include/wine/port.h:252: error: parse error before "pwrite"
../../include/wine/port.h:252: warning: type defaults to `int' in declaration of `pwrite'
../../include/wine/port.h:252: warning: data definition has no type or storage class
make: *** [crypt/crypt.o] Error 1
Good port, we need it for AbiWord 2.2.1,
James
Hello,
I was just looking on how to add FreeDOS to the ReactOS loader,
just as I did for NT bootmenu and Syslinux bootmenu
(both should be working by now)
There's one problem:
I need to update one section "[Operating Systems]" with something like:
FreeDOS="FreeDOS (Beta Distribution)"
and add a section (that's the easy part, and same as for boot.ini),
which I use the following (ASM) program for:
http://www.coli.uni-sb.de/~eric/stuff/soft/specials/text-tool-iniadd.zip
[FreeDOS]
Description="FreeDOS Beta"
BootType=Bootsector
Bootsector=C:\FREEDOS.BSS
I was wondering if you could not remove the [Operating Systems] section
(all it does is provide a label/section description), and do the
following (add description):
[ReactOS]
Description="ReactOS"
BootType=ReactOS
SystemPath=multi(0)disk(0)rdisk(0)partition(1)\ReactOS
[ReactOS_DebugCom1]
"ReactOS (Debug, COM1)"
BootType=ReactOS
SystemPath=multi(0)disk(0)rdisk(0)partition(1)\ReactOS
Options=/DEBUGPORT=COM1
*****instead of:************
[Operating Systems]
ReactOS="ReactOS"
ReactOS_DebugCom1="ReactOS (Debug, COM1)"
[ReactOS]
BootType=ReactOS
SystemPath=multi(0)disk(0)rdisk(0)partition(1)\ReactOS
[ReactOS_DebugCom1]
BootType=ReactOS
SystemPath=multi(0)disk(0)rdisk(0)partition(1)\ReactOS
Options=/DEBUGPORT=COM1
It would require to read the entire freeldr.ini for non-standard
sections ([FREELDR] and [DISPLAY]) like for example [REACTOS]
by the way, sample freeldr.ini does not show how to use Bootsector,
I'll guess I'll just have to install ReactOS over FreeDOS then to figure
this out.
Final request: add a small delay on the LiveCD, so someone can boot to
A: for example from the FreeLDR which is used on cdrom (boot to C: is
already taken care of through cdrom bootsector).
Bernd
Hi!
I was playing around with the output of our build system a bit... I very
much like the way Linux-2.6 does it - it looks like this:
[RC] ntoskrnl.rc
[AS] ke/i386/multiboot.S
[PCH] include/ntoskrnl.h
[CC] ex/i386/interlck.c
[CC] ke/i386/exp.c
...
[CC] inbv/inbv.c
[AR] ntoskrnl.a
[LD] ntoskrnl.nostrip.exe
[RSYM] ntoskrnl.sym
[OBJDUMP] ntoskrnl.map
[LD] ntoskrnl.exe
ntoskrnl.exe was successfully built.
I think everybody will like this output - my question is wether
1) it should become the default (if VERBOSE is unset) and the old output
can be produced if VERBOSE=yes
or
2) the old output should stay the default and the fancy one should be
used if VERBOSE=half or something like this
- blight
winlogin from Windows 2000 and XP creates desktops:
application desktop, winlogon desktop (only winlogon can access this
desktop) and a screen saver desktop.
ReactOS only has one desktop at the moment (correct me if I am false). I
know many security features like login are not implemented. But for a future
release (0.4) ...
Klemens Friedl <frik85>
--
NEU +++ DSL Komplett von GMX +++ http://www.gmx.net/de/go/dsl
GMX DSL-Netzanschluss + Tarif zum superg�nstigen Komplett-Preis!
Gregor Anich schrieb:
>CVSROOT: /CVS/ReactOS
>Module name: reactos
>Repository: reactos/ntoskrnl/include/
>Changes by: blight(a)mok.osexperts.com 04/12/10 06:58:26
>
>Modified files:
> reactos/ntoskrnl/dbg/: print.c
> reactos/ntoskrnl/include/: config.h
>
>Log message:
> Add an option to serialize DbgPrint - don't worry, it's disabled by default and meant only for those who are tired of reading mixed debug messages.
>
>_______________________________________________
>Ros-cvs mailing list
>Ros-cvs(a)reactos.com
>http://reactos.com/mailman/listinfo/ros-cvs
>
>
>
>
Hi,
I like your idea. For my smp build, I've add a simple spinlock which
does serialize the access to the real debug print function. Your
implementation has two problems. The first, it is possible to call
DbgPrint from each interupt level, so it isn't possible to use
ExAllocatePool. The second, we have nice macros like DPRINT, CPRINT or
KEBUGCHECK, which calls DbgPrint for twice. This should be merged in one
call.
- Hartmut
