[hbirr] 14803: Replaced a dirty hack by another.

Replaced a dirty hack by another.
Modified: trunk/reactos/ntoskrnl/ke/main.c
  _____

Modified: trunk/reactos/ntoskrnl/ke/main.c
--- trunk/reactos/ntoskrnl/ke/main.c	2005-04-25 16:17:35 UTC (rev
14802)
+++ trunk/reactos/ntoskrnl/ke/main.c	2005-04-25 17:57:57 UTC (rev
14803)
@@ -103,8 +103,18 @@

MiFreeInitMemory();

/* Never returns */
-        PsIdleThreadMain(NULL);
-        
+#if 0
+	/* FIXME:
+         *   The initial thread isn't a real ETHREAD object, we cannot
call PspExitThread.
+	 */
+	PspExitThread(STATUS_SUCCESS);
+#else
+	while (1) {
+	   LARGE_INTEGER Timeout;
+	   Timeout.QuadPart = 0x7fffffffffffffffLL;
+	   KeDelayExecutionThread(KernelMode, FALSE, &Timeout);
+	}
+#endif        
     } else {

/* Do application processor initialization */

[Alex Ionescu]

[e-mail] wrote:
> Replaced a dirty hack by another.
>
> Modified: trunk/reactos/ntoskrnl/ke/main.c
>
> *Modified: trunk/reactos/ntoskrnl/ke/main.c*
>
> --- trunk/reactos/ntoskrnl/ke/main.c	2005-04-25 16:17:35 UTC (rev 14802)
> +++ trunk/reactos/ntoskrnl/ke/main.c	2005-04-25 17:57:57 UTC (rev 14803)
> @@ -103,8 +103,18 @@
>     MiFreeInitMemory();
>
>     /* Never returns */
>
>
>    PsIdleThreadMain(NULL);
>
> +#if 0
> /* FIXME:
>     *   The initial thread isn't a real ETHREAD object, we cannot call PspExitThread.
> */
> PspExitThread(STATUS_SUCCESS);
>
> +#else
> while (1) {
>   LARGE_INTEGER Timeout;
>   Timeout.QuadPart = 0x7fffffffffffffffLL;
>   KeDelayExecutionThread(KernelMode, FALSE, &Timeout);
> }
>
> +#endif
> } else {
>
>     /* Do application processor initialization */

What's wrong with jumping into the Idle Thread?

It works for me...
But perhaps it doesn't work properly until I can commit the 
Scheduler/Idle thread patch I've been working on...

Best regards,
Alex Ionescu

[Hartmut Birr]

Alex Ionescu wrote:
> [e-mail] wrote:
> > Replaced a dirty hack by another.
> >
> > Modified: trunk/reactos/ntoskrnl/ke/main.c
> >
> > *Modified: trunk/reactos/ntoskrnl/ke/main.c*
> >
> > --- trunk/reactos/ntoskrnl/ke/main.c    2005-04-25 16:17:35 UTC (rev
> > 14802)
> > +++ trunk/reactos/ntoskrnl/ke/main.c    2005-04-25 17:57:57 UTC (rev
> > 14803)
> > @@ -103,8 +103,18 @@
> >     MiFreeInitMemory();
> >             /* Never returns */
> >
> >
> >    PsIdleThreadMain(NULL);
> >
> > +#if 0
> > /* FIXME:
> >     *   The initial thread isn't a real ETHREAD object, we
> >
> > cannot call PspExitThread.
> > */
> > PspExitThread(STATUS_SUCCESS);
> >
> > +#else
> > while (1) {
> >   LARGE_INTEGER Timeout;
> >   Timeout.QuadPart = 0x7fffffffffffffffLL;
> >   KeDelayExecutionThread(KernelMode, FALSE, &Timeout);
> > }
> >
> > +#endif
> > } else {
> >             /* Do application processor initialization */
>
> What's wrong with jumping into the Idle Thread?
>
> It works for me...
> But perhaps it doesn't work properly until I can commit the
> Scheduler/Idle thread patch I've been working on...
>
> Best regards,
> Alex Ionescu
> _______________________________________________
> Ros-diffs mailing list
> [e-mail]
> http://reactos.com:8080/mailman/listinfo/ros-diffs

This thread runs at priority 8. Threads with a lower priority never wake
up. Work items from the delayed work queue are never delivered. If I
compile ros on ros, it needs the double time. So I've delayed the thread
forever.

- Hartmut

[Alex Ionescu]

Hartmut Birr wrote:
> Alex Ionescu wrote:
>
> This thread runs at priority 8. Threads with a lower priority never wake
> up. Work items from the delayed work queue are never delivered. If I
> compile ros on ros, it needs the double time. So I've delayed the thread
> forever.
> Hartmut

You are right, I forgot the way ros initial threads are messed up.
This will work right soon ;)

Best regards,
Alex Ionescu