[Qemu-devel] Re: O/2 Warp 4 and eComStation

[Qemu-devel] Re: O/2 Warp 4 and eComStation

[ecs user]

> From: Natalia Portillo <claunia@claunia.com>
> Try to boot it with less than 64Mb (48Mb).
> OS/2 hangs when 64Mb are informed in the standard
> way they seems to check that in some strange way.
> No matter it is 4.0, 4.5, eCS or whatever.

It does seem like a memory issue.

If I just keep rebooting everytime it hangs, I can
finally get a complete boot. Sometimes I have to
reboot 10-15 times. Once it does boot to the desktop
OS/2 Warp4 runs very nicely. No hangs once completely
booted and very responsive (on 2GHz machine). So you
see it acts more like flaky hardware or flaky driver
(in the emulation).

Need help.


>> --- ecs user <user_ecs@yahoo.com> wrote:
>>
>> There are problems booting Warp4 and eComtation. 
>> System hangs while loading certain drivers (IDE,
>> DASD, Diskette). By repeatedly booting, sometimes
>> can get a successful boot. Might be some type of
>> timing issue.
>  
>> I have some small test images that demonstrate the
>> problem. Are there any developers here familiar
with
>> OS/2 that could look at this issue?
> 
>> Is there a place I can upload the images (15MB)?



		
____________________________________________________ 
Yahoo! Sports 
Rekindle the Rivalries. Sign up for Fantasy Football 
http://football.fantasysports.yahoo.com

Re: [Qemu-devel] Re: O/2 Warp 4 and eComStation

[Tero Kaarlela]

ecs user wrote:

>  
>
>>From: Natalia Portillo <claunia@claunia.com>
>>Try to boot it with less than 64Mb (48Mb).
>>OS/2 hangs when 64Mb are informed in the standard
>>way they seems to check that in some strange way.
>>No matter it is 4.0, 4.5, eCS or whatever.
>>    
>>
>
>It does seem like a memory issue.
>
>If I just keep rebooting everytime it hangs, I can
>finally get a complete boot. Sometimes I have to
>reboot 10-15 times. Once it does boot to the desktop
>OS/2 Warp4 runs very nicely. No hangs once completely
>booted and very responsive (on 2GHz machine). So you
>see it acts more like flaky hardware or flaky driver
>(in the emulation).
>
>Need help.
>

  Well you might start  Qemu with option -d in_asm to see whats 
happening. Then compare log's (the one that freezes and the one that 
works ok) to see whats the difference. That way you might find out where 
the problem is. And you might also take a look at BOCHS emulator page 
because Dimitri Froloff  made few fixes to boot OS/2 under it(some of 
these were memory handling related fixes). He made these fixes lately.


Tero Kaarlela

Re: [Qemu-devel] Re: O/2 Warp 4 and eComStation

[Hampa Hug]

[-- Attachment #1: Type: text/plain, Size: 1116 bytes --]

ecs user wrote:

> > From: Natalia Portillo <claunia@claunia.com>
> > Try to boot it with less than 64Mb (48Mb).
> > OS/2 hangs when 64Mb are informed in the standard
> > way they seems to check that in some strange way.
> > No matter it is 4.0, 4.5, eCS or whatever.
> 
> It does seem like a memory issue.
> 
> If I just keep rebooting everytime it hangs, I can
> finally get a complete boot. Sometimes I have to
> reboot 10-15 times. Once it does boot to the desktop
> OS/2 Warp4 runs very nicely. No hangs once completely
> booted and very responsive (on 2GHz machine). So you
> see it acts more like flaky hardware or flaky driver
> (in the emulation).
> 
> Need help.

I looked into this problem about a year ago. As I
recall, the problem was the PIC generating spurious
interrupts. Specifically, if a pending interrupt is
masked on the PIC, the interrupt request to the CPU
is not reset.

I sent a patch to this list that solved the problem
for me (though it was a hack). Unfortunately I can
no longer reproduce the problem (patch or no patch),
but you might still want to try the attached patch.
YMMV.

Hampa

[-- Attachment #2: qemu-pic-2005-06-17.diff --]
[-- Type: text/plain, Size: 2490 bytes --]

diff -Nur qemu-cvs/hw/i8259.c qemu-local/hw/i8259.c
--- qemu-cvs/hw/i8259.c	2005-06-15 20:09:28.000000000 +0200
+++ qemu-local/hw/i8259.c	2005-06-17 19:12:53.000000000 +0200
@@ -79,20 +79,21 @@
                 s->irr |= mask;
             s->last_irr |= mask;
         } else {
+            s->irr &= ~mask;
             s->last_irr &= ~mask;
         }
     }
 }
 
 /* return the highest priority found in mask (highest = smallest
-   number). Return 8 if no irq */
+   number). Return 16 if no irq */
 static inline int get_priority(PicState *s, int mask)
 {
     int priority;
     if (mask == 0)
-        return 8;
-    priority = 0;
-    while ((mask & (1 << ((priority + s->priority_add) & 7))) == 0)
+        return 16;
+    priority = s->priority_add;
+    while ((mask & (1 << (priority & 7))) == 0)
         priority++;
     return priority;
 }
@@ -104,7 +105,7 @@
 
     mask = s->irr & ~s->imr;
     priority = get_priority(s, mask);
-    if (priority == 8)
+    if (priority == 16)
         return -1;
     /* compute current priority. If special fully nested mode on the
        master, the IRQ coming from the slave is not taken into account
@@ -115,7 +116,7 @@
     cur_priority = get_priority(s, mask);
     if (priority < cur_priority) {
         /* higher priority found: an irq should be generated */
-        return (priority + s->priority_add) & 7;
+        return priority & 7;
     } else {
         return -1;
     }
@@ -132,8 +133,10 @@
     if (irq2 >= 0) {
         /* if irq request by slave pic, signal master PIC */
         pic_set_irq1(&pics[0], 2, 1);
+    } else {
         pic_set_irq1(&pics[0], 2, 0);
     }
+
     /* look at requested irq */
     irq = pic_get_irq(&pics[0]);
     if (irq >= 0) {
@@ -149,6 +152,8 @@
         printf("pic: cpu_interrupt\n");
 #endif
         cpu_interrupt(cpu_single_env, CPU_INTERRUPT_HARD);
+    } else {
+        cpu_reset_interrupt(cpu_single_env, CPU_INTERRUPT_HARD);
     }
 }
 
@@ -297,8 +304,8 @@
             case 1: /* end of interrupt */
             case 5:
                 priority = get_priority(s, s->isr);
-                if (priority != 8) {
-                    irq = (priority + s->priority_add) & 7;
+                if (priority != 16) {
+                    irq = priority & 7;
                     s->isr &= ~(1 << irq);
                     if (cmd == 5)
                         s->priority_add = (irq + 1) & 7;

[Qemu-devel] Re: O/2 Warp 4 and eComStation

[ecs user]

Sorry, there was a MAJOR typo in my last message. It
was supposed to say
"It does NOT seem like a memory issue."

Wouldn't it be more a BIOS or (emulated) hardware
timing issue.

The hangs typically happen at the time of laoding
certain hardware drivers
IDE
diskette
etc

> -- ecs user <user_ecs@yahoo.com> wrote:
> 
> It does seem like a memory issue.
> 
> If I just keep rebooting everytime it hangs, I can
> finally get a complete boot. Sometimes I have to
> reboot 10-15 times. Once it does boot to the desktop
> OS/2 Warp4 runs very nicely. No hangs once
> completely
> booted and very responsive (on 2GHz machine). So you
> see it acts more like flaky hardware or flaky driver
> (in the emulation).
> 
> Need help.



		
____________________________________________________ 
Yahoo! Sports 
Rekindle the Rivalries. Sign up for Fantasy Football 
http://football.fantasysports.yahoo.com

[Qemu-devel] Re: O/2 Warp 4 and eComStation

[ecs user]

>  Tero Kaarlela wrote:
>  Well you might start Qemu with option -d in_asm

I do not see that "-d in_asm" option for running QEMU.

Is "-d in_asm" a run option or is a compile of qemu
option.


> Then compare log's (the one that freezes and the one
> that works ok) to see whats the difference.
>  That way you might find out where the problem is.



		
____________________________________________________ 
Yahoo! Sports 
Rekindle the Rivalries. Sign up for Fantasy Football 
http://football.fantasysports.yahoo.com

Re: [Qemu-devel] Re: O/2 Warp 4 and eComStation

[Tero Kaarlela]

ecs user wrote:

>> Tero Kaarlela wrote:
>> Well you might start Qemu with option -d in_asm
>>    
>>
>
>I do not see that "-d in_asm" option for running QEMU.
>
>Is "-d in_asm" a run option or is a compile of qemu
>option.
>

   From Qemu manual  section 3.3 Invocation:


   `-d'
Output log in /tmp/qemu.log

  Runnig option ie : qemu -m 64 -d in_asm -hda os2.img



Tero Kaarlela

[Qemu-devel] Re: O/2 Warp 4 and eComStation

[ecs user]

Hampa,

Thanks for the tip. I will try your patch.

What versions of OS/2 do you run?

> Hampa Hug <hampa@hampa.ch> wrote:
> I looked into this problem about a year ago.
> As I recall, the problem was the PIC generating
> spurious interrupts. Specifically, if a pending
> interrupt is masked on the PIC, the interrupt
> request to the CPU is not reset.

> I sent a patch to this list that solved the problem
> for me (though it was a hack). Unfortunately I can
> no longer reproduce the problem (patch or no patch),
> but you might still want to try the attached patch.
 


	
		
__________________________________ 
Do you Yahoo!? 
Yahoo! Mail - You care about security. So do we. 
http://promotions.yahoo.com/new_mail