reactos-kernel

[ros-kernel] compile error

[David K. <kr...@mb...>]

make -C ntoskrnl
make[1]: Entering directory `C:/tmp/reactos/ntoskrnl'
gcc  -I./include -pipe -march=i386 -I./include -I./include -D__NTOSKRNL__ -g -Wall -Werror -I../include -pipe -marc
h=i386 -c ldr/loader.c -o ldr/loader.o
ldr/loader.c: In function `LdrInitializeBootStartDriver':
ldr/loader.c:1075: redeclaration of `Length'
ldr/loader.c:1064: `Length' previously declared here
ldr/loader.c:1076: conflicting types for `Ext'
ldr/loader.c:1066: previous declaration of `Ext'
mingwlibgcc-libmingw322.95.3-7cc1.exe: warnings being treated as errors
ldr/loader.c:1114: warning: implicit declaration of function `LdrFindModuleObject'
ldr/loader.c:1168: warning: assignment from incompatible pointer type
ldr/loader.c:1170: invalid operands to binary -
make[1]: *** [ldr/loader.o] Error 1
make[1]: Leaving directory `C:/tmp/reactos/ntoskrnl'
make: *** [ntoskrnl] Error 2

David

RE: [ros-kernel] Patch: Teach freeldr registry aboutmultiandexpanded types

[Kohn E. D. <em...@cs...>]

On Thu, 13 Jun 2002, Aliberti Emanuele wrote:

> >The ability to cast arrays to pointers without using & is a compiler
> >trick. A dangerous trick IMHO.
>
> It's not casting. Arrays ARE pointers in C. Square brackets are a language construct to avoid explicit pointer computations.

Arrays are *not* pointers in C. However, with few exceptions (such as
passing them as arguments to sizeof()), an array is similar to a pointer
to its first element.
...
int arr[10];
int* pi = arr;
...

The expression &arr will yield something of type `pointer to array of 10
integers'. While both arr and &arr yield the same value when printed,
these pointers are *different types*, in that, for example, arr + 1 will
yield the address of the second element in the array (i.e. &arr[1]), while
&arr + 1 will yield the address of one element past the end of the array
(i.e.  &arr[10]). There are of course other differences, too.

Obviously pointers to integers are not compatible with pointers to arrays.
Therefore the following code will not compile without an error or at least
a warning from the compiler:

...
int arr[10];
int* pi = &arr;
...

To designate a pointer to array you need to use the following syntax:
int (*parr)[10] = &arr;

I hope that this clarifies the issues.


							Emil

>
> _______________________________________________________________
>
> Don't miss the 2002 Sprint PCS Application Developer's Conference
> August 25-28 in Las Vegas - http://devcon.sprintpcs.com/adp/index.cfm?source=osdntextlink
>
> _______________________________________________
> reactos-kernel mailing list
> rea...@li...
> https://lists.sourceforge.net/lists/listinfo/reactos-kernel
>

[ros-kernel] CVS Commit: reactos

[Eric K. <ek...@rz...>]

system.hiv
    Added VGA driver.

subsys/csrss/init.c
drivers/dd/vidport/vidport.c
    Pass pointer to CSRSS process on first create request, which is issued
by CSRSS.

subsys/win32k/eng/device.c
    Fixed a direct reference to the miniport driver object.

These changes make gditest run again. :-)


Eric

Re: [ros-kernel] DocBook authoring tools

[Robert K. <ro...@ko...>]

In prevention of what might happen, I write:
_Stop_ the arising editor flame.


KJK::Hyperion schrieb:

> At 22.52 12/06/2002, you wrote:
> >They are free for personal use and professional use.
> >They work almost out of the box.
>
> "almost"? ;-(
>
> >Note: they are Java applications, therefore you need at least a 400 MHz
> >processor with 256 Mbyte RAM.
>
> ack! personally, I think Java apps suck, and deserve a quick, silent
> death... what about native applications? on Windows a supposedly cool
> program is XML Cooktop - but in fact it sucks. May I suggest SciTE? works
> on Windows and Linux, it's not specifically a XML editor but a generic code
> editor, like ViM. Unlike ViM, it's usable. But, unlike "real" XML editors,
> it's not DTD-aware, i.e. no pretty property boxes, you have to do it the
> hard way, by hand. Good thing is that you can also use it as C/C++ editor
> (in fact, I did all my coding for ReactOS with SciTE), but it supports a
> ton of other languages as well (Ruby, Python, Pascal, Ada, Perl, OLE IDL,
> PHP, SQL, PLSQL, resource scripts, TCL, VB, etc.). Plus, no crazy memory or
> CPU requirements, a Pentium 133 with 16 MB RAM will do
>
> check it out:
> <http://www.scintilla.org/SciTE.html>
>
> (the default config, as usual, works only for the developer. Before using
> it, read the manual, or you may miss some cool features)
>
> _______________________________________________________________
>
> Don't miss the 2002 Sprint PCS Application Developer's Conference
> August 25-28 in Las Vegas - http://devcon.sprintpcs.com/adp/index.cfm?source=osdntextlink
>
> _______________________________________________
> reactos-kernel mailing list
> rea...@li...
> https://lists.sourceforge.net/lists/listinfo/reactos-kernel

Re: [ros-kernel] Can't run winhello.exe

[Eric K. <ek...@rz...>]

"Jason Filby" <jas...@ya...> wrote:

> >Sorry, that was a typo. I meant vgamp.sys. Vidport.sys should
> >handle the create requests and retrieve the process id or handle
> from
> >csrss and do the initialization for int10 calls.
>
> But doesn't it already do this?
It does, but now vidport.sys is loaded when vgamp.sys is loaded. Both are
system-start drivers. Csrss does not run at that time, so opening the csrss
process by its id will fail! :-/

> >Btw, I guess it is useful to rename vidport.sys to videoprt.sys
> >since other video drivers will rely on the name of the port driver.
>
> I thought other drivers would look for the device name in the object
> manager?
Vidport.sys does not appear in the object managers name space. It is now
automatcially loaded by ntoskrnl when vgamp.sys is relocated and no loaded
'vidport.sys' module/driver can be found. This is very much like the
automatic loading of dlls. Btw, the same will apply to scsiport.sys and
class2.sys, or better 'all port drivers'.

Eric

Re: [ros-kernel] Can't run winhello.exe

[Jason F. <jas...@ya...>]

--- Eric Kohl <ek...@rz...> wrote:
>Sorry, that was a typo. I meant vgamp.sys. Vidport.sys should
>handle the create requests and retrieve the process id or handle
from
>csrss and do the initialization for int10 calls.

But doesn't it already do this?

>Btw, I guess it is useful to rename vidport.sys to videoprt.sys
>since other video drivers will rely on the name of the port driver.

I thought other drivers would look for the device name in the object
manager?

- Jason








__________________________________________________
Do You Yahoo!?
Yahoo! - Official partner of 2002 FIFA World Cup
http://fifaworldcup.yahoo.com

Re: [ros-kernel] Can't run winhello.exe

[Eric K. <ek...@rz...>]

"Jason Filby" <jas...@ya...> wrote:

> AFAIK Vgaddi does not ever an int10 bios call - only Vgamp does this.
> Vgaddi makes calls to Vgamp if it needs something like video
> initialization done (which issues an int10 call).

Sorry, that was a typo. I meant vgamp.sys. Vidport.sys should handle the
create requests and retrieve the process id or handle from csrss and do the
initialization for int10 calls.

Btw, I guess it is useful to rename vidport.sys to videoprt.sys since other
video drivers will rely on the name of the port driver.

Eric

Re: [ros-kernel] Can't run winhello.exe

[Jason F. <jas...@ya...>]

Hey Eric!

>>Is that code committed, or do you mean just added to your own
>>tree?
>Now it's in the cvs tree.

Great :)

>There is still one problem left. Vgaddi needs the handle of the
>csrss process to issue int10 bios calls. The current method of
>retrieving this handle will not work. I suggest csrss should open
and
>close all \\Device\\VideoX devices upon its initialization. Then
>vidport can retrieve the callers process id or process handle upon
the
>first open request.

AFAIK Vgaddi does not ever an int10 bios call - only Vgamp does this.
Vgaddi makes calls to Vgamp if it needs something like video
initialization done (which issues an int10 call).

- Jason








__________________________________________________
Do You Yahoo!?
Yahoo! - Official partner of 2002 FIFA World Cup
http://fifaworldcup.yahoo.com

Re: [ros-kernel] Can't run winhello.exe

[Eric K. <ek...@rz...>]

"Jason Filby" <jas...@ya...> wrote:

Good morning Jason!

> Is that code committed, or do you mean just added to your own tree?

Now it's in the cvs tree.

> I'm not sure about right now... but it will surely have to load
> vgamp.sys in the future (when we get it working properly).

Okay, I fixed the vgamp/vidport mess. At least I think so. ;-)
There is still one problem left. Vgaddi needs the handle of the csrss
process to issue int10 bios calls. The current method of retrieving this
handle will not work. I suggest csrss should open and close all
\\Device\\VideoX devices upon its initialization. Then vidport can retrieve
the callers process id or process handle upon the first open request.


Eric

[ros-kernel] CVS Commit: reactos

[Eric K. <ek...@rz...>]

include/ddk/zwtypes.h
ntoskrnl/ex/sysinfo.c
ntoskrnl/include/internal/ldr.h
ntoskrnl/ldr/loader.c
subsys/smss/init.c
subsys/win32k/ldr/loader.c
subsys/win32k/main/dllmain.c
subsys/win32k/misc/driver.c
    Fixed automatic loading of required modules.
    Cleaned up module loading interface.


Eric

Re: [ros-kernel] Can't run winhello.exe

[Jason F. <jas...@ya...>]

--- Eric Kohl <ek...@rz...> wrote:
>Yes, it can be a regression from the driver changes!
>I have seen that the vgamp/vidport loading sequence is pretty
>braindead and surely broken due to the latest changes.
>I already added some code to LdrProcessPEModule() that auto-loads
>modules that are not currently loaded, but used by the currenly
>loading driver. So when vgamp.sys is loaded, the module loader will
>automatically load vidport.sys.

Is that code committed, or do you mean just added to your own tree?

>I will also have to examine the code in win32k.sys that loads the
>display driver (vgaddi.sys). This could also be broken.
>Do you load vgamp.sys or just vidport.sys to run winhello.exe?

I'm not sure about right now... but it will surely have to load
vgamp.sys in the future (when we get it working properly).

- Jason








__________________________________________________
Do You Yahoo!?
Yahoo! - Official partner of 2002 FIFA World Cup
http://fifaworldcup.yahoo.com

Re: [ros-kernel] Bootloader and Kernel

[James M. <jid...@sa...>]

Okay thanks.

James Marjie
GnuPG KeyID:    0x7C837497

"Take your life in your own hands, and what happens? A terrible thing: no
one to blame." -Erica Jong

Re: [ros-kernel] Bootloader and Kernel

[David W. <we...@cw...>]

On Thu, Jun 13, 2002 at 01:41:43PM -0700, James Marjie wrote:
> I was interested in knowing if reactos will be multiboot compliant? So 
> that you could eliminate the need for chain loading when attempting to 
> load from grub or lilo( This would be some point in the future.).
> 
ReactOS is multiboot compliant. You should be able to boot directly
from grub.

Re: [ros-kernel] NtDuplicateObject

[David W. <we...@cw...>]

On Thu, Jun 13, 2002 at 06:26:21PM +0200, Hartmut Birr wrote:
> While implementing of c-runtime-piping, I had some problems with duplicated 
> handles. I'm not sure, if the problem was in NtDuplicateObject or in 
> ObCreateHandleTable.
> 
Could you give some more details; I'd like to sort this out.

Re: [ros-kernel] fprintf brokeness?

[James M. <jid...@sa...>]

Okay will do. I should get the latest source tree from CVS probaly then.

James Marjie
GnuPG KeyID:    0x7C837497

"Take your life in your own hands, and what happens? A terrible thing: no
one to blame." -Erica Jong

Re: [ros-kernel] wine dll's on a real system.

[Diego I. <ias...@ac...>]

On Thursday 13 June 2002 05:28, Steven Edwards wrote:
> Thanks for your time in testing Diego, lets hope next time WINE will be
> a little more stable for you =). If you are still interested in doing
> some testing I will keep you posted on my progress with the wine port. I
> have taken the past week off to focus on a few other things but am
> forwarding your report to wine-devel and will be resuming work this
> weekend.
I will look for replies. I am here if you want to continue the testings.

> Thanks Again
> Steven
no problems (exept we got kicked of this stupid mudial....)

 - diego

-- 
What PROGRAM are they watching?

[ros-kernel] Bootloader and Kernel

[James M. <jid...@sa...>]

Hi,
    I was interested in knowing if reactos will be multiboot compliant? =
So that you could eliminate the need for chain loading when attempting =
to load from grub or lilo( This would be some point in the future.).

James Marjie
GnuPG KeyID:    0x7C837497

"Take your life in your own hands, and what happens? A terrible thing: =
no one to blame." -Erica Jong

Re: [ros-kernel] fprintf brokeness?

[Steven E. <ste...@ya...>]

Its working now. *fprintf is good, but you can look at
cscanf I think. There is a bug report on the message
board on reactos.com. Also you can take a look at
rosapps/mc. Midnight Commander for Win32 has a bug
when changing directorys under ReactOS.

Thanks
Steven


--- James Marjie <jid...@sa...> wrote:
> Okay, might have take a look and see if I can help.
> 
> James Marjie
> GnuPG KeyID:    0x7C837497
> 
> "Take your life in your own hands, and what happens?
> A terrible thing: no
> one to blame." -Erica Jong


__________________________________________________
Do You Yahoo!?
Yahoo! - Official partner of 2002 FIFA World Cup
http://fifaworldcup.yahoo.com

Re: [ros-kernel] fprintf brokeness?

[James M. <jid...@sa...>]

Okay, might have take a look and see if I can help.

James Marjie
GnuPG KeyID:    0x7C837497

"Take your life in your own hands, and what happens? A terrible thing: no
one to blame." -Erica Jong

Re: [ros-kernel] Can't run winhello.exe

[Eric K. <ek...@rz...>]

"Jason Filby" <jas...@ya...> wrote:

> Winhello.exe can't run anymore. It used to get to debug print
> information, but now it just sits there -- not even a crash message.
> Is this a possible regression from all the driver changes?

Yes, it can be a regression from the driver changes!
I have seen that the vgamp/vidport loading sequence is pretty braindead and
surely broken due to the latest changes. I already added some code to
LdrProcessPEModule() that auto-loads modules that are not currently loaded,
but used by the currenly loading driver. So when vgamp.sys is loaded, the
module loader will automatically load vidport.sys.

I will also have to examine the code in win32k.sys that loads the display
driver (vgaddi.sys). This could also be broken.

Do you load vgamp.sys or just vidport.sys to run winhello.exe?


Eric

[ros-kernel] Can't run winhello.exe

[Jason F. <jas...@ya...>]

Hey all

Winhello.exe can't run anymore. It used to get to debug print
information, but now it just sits there -- not even a crash message.
Is this a possible regression from all the driver changes?

Thanks
- Jason







__________________________________________________
Do You Yahoo!?
Yahoo! - Official partner of 2002 FIFA World Cup
http://fifaworldcup.yahoo.com

[ros-kernel] Sscanf question

[Steven E. <Ste...@ya...>]

I don't know if anyone addressed this but there was no followup in the
user support board on ReactOS.com

Author: gl (---.irtel.ru)
Date:   06-02-02 03:24

It seems that sscanf is not correctly done.Am I right?
This program hangs on sscanf:
----
int main(int argc, char *argv[]){
int i;
if(argc>1) {
sscanf(argv[1], "%i", &i);
printf("%i\n", i);
return 0;
};
};

Thanks
Steven

"Every revolution was once a thought in one man's mind"
- Ralph Waldo Emerson 

RE: [ros-kernel] CVS Commit: reactos

[Steven E. <Ste...@ya...>]

> lib\msvcrt\vfwprint.c:
> lib\msvcrt\vfprintf.c:
>   Reseting the conversion qualifier for each new value from 
> the format string.

Thanks Hartmut. I will test it in just a sec.
Steven

"Every revolution was once a thought in one man's mind"
- Ralph Waldo Emerson 

RE: [ros-kernel] QuerySystemInformation mini patch

[Steven E. <Ste...@ya...>]

I didn't see anyone comment on this, is anyone working on it? This is
needed with wine for psapi but no one on the WINE side seems interested
in impleneting it right noe.

Thanks
Steven

"Every revolution was once a thought in one man's mind"
- Ralph Waldo Emerson 

> -----Original Message-----
> From: rea...@li... 
> [mailto:rea...@li...] On Behalf 
> Of cra...@mi...
> Sent: Wednesday, June 12, 2002 11:51 AM
> To: rea...@li...
> Subject: [ros-kernel] QuerySystemInformation mini patch
> 
> 
> see the attached file.
> I would like to know if someone is already working on this API ?
> 
> last thing:
> QSI_DEF(SystemPathInformation)
> {
> 	/* FIXME: QSI returns STATUS_BREAKPOINT. Why? */
> 	return (STATUS_BREAKPOINT);
> }
> ==> DbgPrint("EX: SystemPathInformation now available via 
> SharedUserData\n") here is the answer.
> 
> -- crazylord
> 

[ros-kernel] CVS Commit: reactos

[Hartmut B. <har...@te...>]

lib\msvcrt\vfwprint.c:
lib\msvcrt\vfprintf.c:
  Reseting the conversion qualifier for each new value from the format string.

- Hartmut