Re:[Qemu-devel] Interrupt request info

Re:[Qemu-devel] Interrupt request info

[Roger Lathrop]

Alessandro,

All you should need to do to raise an IRQ in your code is:
pic_set_irq(s->irq,1);

When the irq is serviced (in one of your ioport traps, I would assume), 
knock the irq down:
pic_set_irq(s->irq,0);

For debugging, you might want to go into i8259.c and uncomment the #define 
DEBUG_IRQ_COUNT line.
You can then use command 'info irq' in the monitor console. This will show 
you a count of interrupts for each IRQ.

Regards,
Roger

> For example, I wrote a module of a simple memory without implementing 
> IRQ...
> I think it is  an error but it works fine... are IRQ raise in any case???
> I look at other code source of implemented hw but for each device there is 
> a
> differeny way to raise an interrupt (i see).
> Is there a standard way to do this?
> The following is an excerpt of my simple memory, please can you describe 
> me
> fastly how can I manage this?
>
> Thank you

Re: [Qemu-devel] Interrupt request info

[Paul Brook]

On Wednesday 01 November 2006 23:25, Roger Lathrop wrote:
> Alessandro,
>
> All you should need to do to raise an IRQ in your code is:
> pic_set_irq(s->irq,1);
>
> When the irq is serviced (in one of your ioport traps, I would assume),
> knock the irq down:
> pic_set_irq(s->irq,0);

No. You should use pic_set_irq_new. Otherwise your code will break on machines 
with multiple interrupt controllers.

Paul

Re: [Qemu-devel] Interrupt request info

[Alessandro Corradi]

[-- Attachment #1: Type: text/plain, Size: 1453 bytes --]

Thank you very much for the answer,
so the only think to do to raise correctly an interrupt (I think) is only to
raise it when a read funciton is completed (in case of my simple memory)
because it will alarm the process in waiting status to read the value in the
register.

the workaround of a read operation of a cell of this memory is:
1)write address of the cell that I want to read in the address register,
2)send the execution request to get the value of the cell indicated by the
address reg (write on the command register the read opcode),
3)wait for control register response untill read operation is executed.
4)read the data register.

so I think that interrupt request must be between point 3 and 4.
Do you agree with my consideration?

I hope that writed something understandable :)

Thanks

Ale


2006/11/2, Paul Brook <paul@codesourcery.com>:
>
> On Wednesday 01 November 2006 23:25, Roger Lathrop wrote:
> > Alessandro,
> >
> > All you should need to do to raise an IRQ in your code is:
> > pic_set_irq(s->irq,1);
> >
> > When the irq is serviced (in one of your ioport traps, I would assume),
> > knock the irq down:
> > pic_set_irq(s->irq,0);
>
> No. You should use pic_set_irq_new. Otherwise your code will break on
> machines
> with multiple interrupt controllers.
>
> Paul
>
>
> _______________________________________________
> Qemu-devel mailing list
> Qemu-devel@nongnu.org
> http://lists.nongnu.org/mailman/listinfo/qemu-devel
>

[-- Attachment #2: Type: text/html, Size: 1913 bytes --]

Fwd: [Qemu-devel] Interrupt request info

[Alessandro Corradi]

[-- Attachment #1: Type: text/plain, Size: 1641 bytes --]

---------- Forwarded message ----------
From: Alessandro Corradi <ale.corradi@gmail.com>
Date: 4-nov-2006 17.08
Subject: Re: [Qemu-devel] Interrupt request info
To: qemu-devel@nongnu.org

Thank you very much for the answer,
so the only think to do to raise correctly an interrupt (I think) is only to
raise it when a read funciton is completed (in case of my simple memory)
because it will alarm the process in waiting status to read the value in the
register.

the workaround of a read operation of a cell of this memory is:
1)write address of the cell that I want to read in the address register,
2)send the execution request to get the value of the cell indicated by the
address reg (write on the command register the read opcode),
3)wait for control register response untill read operation is executed.
4)read the data register.

so I think that interrupt request must be between point 3 and 4.
Do you agree with my consideration?

I hope that writed something understandable :)

Thanks

Ale


2006/11/2, Paul Brook <paul@codesourcery.com>:
>
> On Wednesday 01 November 2006 23:25, Roger Lathrop wrote:
> > Alessandro,
> >
> > All you should need to do to raise an IRQ in your code is:
> > pic_set_irq(s->irq,1);
> >
> > When the irq is serviced (in one of your ioport traps, I would assume),
> > knock the irq down:
> > pic_set_irq(s->irq,0);
>
> No. You should use pic_set_irq_new. Otherwise your code will break on
> machines
> with multiple interrupt controllers.
>
> Paul
>
>
> _______________________________________________
> Qemu-devel mailing list
> Qemu-devel@nongnu.org
> http://lists.nongnu.org/mailman/listinfo/qemu-devel
>

[-- Attachment #2: Type: text/html, Size: 2563 bytes --]

Re: [Qemu-devel] Interrupt request info

[Alessandro Corradi]

[-- Attachment #1: Type: text/plain, Size: 1000 bytes --]

Thank you very much,
But I don't understand what the first parameter is on pic_set_irq_new...
I tried anyway to implement my hw with pic_set_irq, but in qemu console I
didn't see my irq (13) with info irq.
I tried to raise it: pic_set_irq(s->irq,1);

What I need to do to raise irq correctly with pic_set_irq_new specially the
first parameter)?

Thank you

2006/11/2, Paul Brook <paul@codesourcery.com>:
>
> On Wednesday 01 November 2006 23:25, Roger Lathrop wrote:
> > Alessandro,
> >
> > All you should need to do to raise an IRQ in your code is:
> > pic_set_irq(s->irq,1);
> >
> > When the irq is serviced (in one of your ioport traps, I would assume),
> > knock the irq down:
> > pic_set_irq(s->irq,0);
>
> No. You should use pic_set_irq_new. Otherwise your code will break on
> machines
> with multiple interrupt controllers.
>
> Paul
>
>
> _______________________________________________
> Qemu-devel mailing list
> Qemu-devel@nongnu.org
> http://lists.nongnu.org/mailman/listinfo/qemu-devel
>

[-- Attachment #2: Type: text/html, Size: 1437 bytes --]

Re: [Qemu-devel] Interrupt request info

[Alessandro Corradi]

[-- Attachment #1: Type: text/plain, Size: 1032 bytes --]

I tried with both pic_set_irq and pic_set_irq_new but nothing happen (the
irq number is 13 and I verified with info irq in qemu monitor).
I tried also to hack i8259.c and the function is not called (I think).
What the first parameter of pic_set_irq_new is referred to?

Another question: why a new virtual device works anyway without interrupt
mechanism? isn't it essential?

Thanks

Ale

2006/11/2, Paul Brook <paul@codesourcery.com>:
>
> On Wednesday 01 November 2006 23:25, Roger Lathrop wrote:
> > Alessandro,
> >
> > All you should need to do to raise an IRQ in your code is:
> > pic_set_irq(s->irq,1);
> >
> > When the irq is serviced (in one of your ioport traps, I would assume),
> > knock the irq down:
> > pic_set_irq(s->irq,0);
>
> No. You should use pic_set_irq_new. Otherwise your code will break on
> machines
> with multiple interrupt controllers.
>
> Paul
>
>
> _______________________________________________
> Qemu-devel mailing list
> Qemu-devel@nongnu.org
> http://lists.nongnu.org/mailman/listinfo/qemu-devel
>

[-- Attachment #2: Type: text/html, Size: 1470 bytes --]