[U-Boot-Users] question about interrupt

[U-Boot-Users] question about interrupt

[loody]

Dear all:
I have trace interrupt/Fiq vector in uboot based on smdk2410.
after we save user registers, we jump to following functions I
excerpted at the end of mail.
But these 2 functions seems not really handle the irq/fiq, would
someone please tell me where the real
interrupt handler is?

Is the interrupt handled in kernel?
if it really is, how loader pass the vector pointer to OS?
appreciate your help,
miloody


void do_fiq (struct pt_regs *pt_regs)
{
	printf ("fast interrupt request\n");
	show_regs (pt_regs);
	bad_mode ();
}

void do_irq (struct pt_regs *pt_regs)
{
#if defined (CONFIG_USE_IRQ) && defined (CONFIG_ARCH_INTEGRATOR)
	/* ASSUMED to be a timer interrupt  */
	/* Just clear it - count handled in */
	/* integratorap.c                   */
	*(volatile ulong *)(CFG_TIMERBASE + 0x0C) = 0;
#else
	printf ("interrupt request\n");
	show_regs (pt_regs);
	bad_mode ();
#endif
}

[U-Boot-Users] question about interrupt

[Chetan Nanda]

On Tue, May 20, 2008 at 7:41 AM, loody <miloody@gmail.com> wrote:
> Dear all:
> I have trace interrupt/Fiq vector in uboot based on smdk2410.
> after we save user registers, we jump to following functions I
> excerpted at the end of mail.
> But these 2 functions seems not really handle the irq/fiq, would
> someone please tell me where the real
> interrupt handler is?
>
> Is the interrupt handled in kernel?
> if it really is, how loader pass the vector pointer to OS?
> appreciate your help,
> miloody
>
>
> void do_fiq (struct pt_regs *pt_regs)
> {
>        printf ("fast interrupt request\n");
>        show_regs (pt_regs);
>        bad_mode ();
> }
>
> void do_irq (struct pt_regs *pt_regs)
> {
> #if defined (CONFIG_USE_IRQ) && defined (CONFIG_ARCH_INTEGRATOR)
>        /* ASSUMED to be a timer interrupt  */
>        /* Just clear it - count handled in */
>        /* integratorap.c                   */
>        *(volatile ulong *)(CFG_TIMERBASE + 0x0C) = 0;
> #else
>        printf ("interrupt request\n");
>        show_regs (pt_regs);
>        bad_mode ();
> #endif
> }
>
As far as i know, U-Boot do no support interrupts, and above functions
are just the place holders.
and also Interrupts for various devices are generally not enabled in
u-boot code.

> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2008.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> U-Boot-Users mailing list
> U-Boot-Users at lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/u-boot-users
>

[U-Boot-Users] question about interrupt

[loody]

hi:
thanks for your kind help.
As far as I know, the cpu will set the PC to irq/fiq vector when the
these exceptions are triggered. And these pointers are usually at the
beginning position of bootloader, if bootloader wants to pass these
functions to kernel, there may be some instructions like "b xxxxxx'.

If uboot don't handle these exceptions, how the kernel knows these events?
appreciate your kind help,
miloody

2008/5/20 Chetan Nanda <chetannanda@gmail.com>:
> On Tue, May 20, 2008 at 7:41 AM, loody <miloody@gmail.com> wrote:
>> Dear all:
>> I have trace interrupt/Fiq vector in uboot based on smdk2410.
>> after we save user registers, we jump to following functions I
>> excerpted at the end of mail.
>> But these 2 functions seems not really handle the irq/fiq, would
>> someone please tell me where the real
>> interrupt handler is?
>>
>> Is the interrupt handled in kernel?
>> if it really is, how loader pass the vector pointer to OS?
>> appreciate your help,
>> miloody
>>
>>
>> void do_fiq (struct pt_regs *pt_regs)
>> {
>>        printf ("fast interrupt request\n");
>>        show_regs (pt_regs);
>>        bad_mode ();
>> }
>>
>> void do_irq (struct pt_regs *pt_regs)
>> {
>> #if defined (CONFIG_USE_IRQ) && defined (CONFIG_ARCH_INTEGRATOR)
>>        /* ASSUMED to be a timer interrupt  */
>>        /* Just clear it - count handled in */
>>        /* integratorap.c                   */
>>        *(volatile ulong *)(CFG_TIMERBASE + 0x0C) = 0;
>> #else
>>        printf ("interrupt request\n");
>>        show_regs (pt_regs);
>>        bad_mode ();
>> #endif
>> }
>>
> As far as i know, U-Boot do no support interrupts, and above functions
> are just the place holders.
> and also Interrupts for various devices are generally not enabled in
> u-boot code.
>
>> -------------------------------------------------------------------------
>> This SF.net email is sponsored by: Microsoft
>> Defy all challenges. Microsoft(R) Visual Studio 2008.
>> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
>> _______________________________________________
>> U-Boot-Users mailing list
>> U-Boot-Users at lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/u-boot-users
>>
>

[U-Boot-Users] question about interrupt

[Chetan Nanda]

On Tue, May 20, 2008 at 11:23 AM, loody <miloody@gmail.com> wrote:
> hi:
> thanks for your kind help.
> As far as I know, the cpu will set the PC to irq/fiq vector when the
> these exceptions are triggered. And these pointers are usually at the
> beginning position of bootloader, if bootloader wants to pass these
> functions to kernel, there may be some instructions like "b xxxxxx'.
>
> If uboot don't handle these exceptions, how the kernel knows these events?
> appreciate your kind help,
> miloody
>
as i have mentioned, u-boot do not handle any interrupt, and neither
it pass vector information to OS it going to boot.
But OS that booted specify its own vector tables for handling the
interrupts and exception.

Anyone, correct me if i am at wrong path.


> 2008/5/20 Chetan Nanda <chetannanda@gmail.com>:
>> On Tue, May 20, 2008 at 7:41 AM, loody <miloody@gmail.com> wrote:
>>> Dear all:
>>> I have trace interrupt/Fiq vector in uboot based on smdk2410.
>>> after we save user registers, we jump to following functions I
>>> excerpted at the end of mail.
>>> But these 2 functions seems not really handle the irq/fiq, would
>>> someone please tell me where the real
>>> interrupt handler is?
>>>
>>> Is the interrupt handled in kernel?
>>> if it really is, how loader pass the vector pointer to OS?
>>> appreciate your help,
>>> miloody
>>>
>>>
>>> void do_fiq (struct pt_regs *pt_regs)
>>> {
>>>        printf ("fast interrupt request\n");
>>>        show_regs (pt_regs);
>>>        bad_mode ();
>>> }
>>>
>>> void do_irq (struct pt_regs *pt_regs)
>>> {
>>> #if defined (CONFIG_USE_IRQ) && defined (CONFIG_ARCH_INTEGRATOR)
>>>        /* ASSUMED to be a timer interrupt  */
>>>        /* Just clear it - count handled in */
>>>        /* integratorap.c                   */
>>>        *(volatile ulong *)(CFG_TIMERBASE + 0x0C) = 0;
>>> #else
>>>        printf ("interrupt request\n");
>>>        show_regs (pt_regs);
>>>        bad_mode ();
>>> #endif
>>> }
>>>
>> As far as i know, U-Boot do no support interrupts, and above functions
>> are just the place holders.
>> and also Interrupts for various devices are generally not enabled in
>> u-boot code.
>>
>>> -------------------------------------------------------------------------
>>> This SF.net email is sponsored by: Microsoft
>>> Defy all challenges. Microsoft(R) Visual Studio 2008.
>>> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
>>> _______________________________________________
>>> U-Boot-Users mailing list
>>> U-Boot-Users at lists.sourceforge.net
>>> https://lists.sourceforge.net/lists/listinfo/u-boot-users
>>>
>>
>

[U-Boot-Users] question about interrupt

[Wolfgang Denk]

In message <7f245da80805200136y297e5309k110f1b193981ac16@mail.gmail.com> you wrote:
>
> as i have mentioned, u-boot do not handle any interrupt, and neither

In this general form, the statement is wrong.

Depending on architecture, U-Boot does implement and use  interrupts.
It  even exports an interface to standalone applications which allows
to register custom interrupt handlers.

It's just just most U-Boot drivers do not use interrupts but polling
mode - note that there are exceptions from this rule, too.

Best regards,

Wolfgang Denk

-- 
DENX Software Engineering GmbH,     MD: Wolfgang Denk & Detlev Zundel
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd at denx.de
A modem is a baudy house.

[U-Boot-Users] question about interrupt

[Chetan Nanda]

On Tue, May 20, 2008 at 2:53 PM, Wolfgang Denk <wd@denx.de> wrote:
> In message <7f245da80805200136y297e5309k110f1b193981ac16@mail.gmail.com> you wrote:
>>
>> as i have mentioned, u-boot do not handle any interrupt, and neither
>
> In this general form, the statement is wrong.
>
> Depending on architecture, U-Boot does implement and use  interrupts.
> It  even exports an interface to standalone applications which allows
> to register custom interrupt handlers.

So if someone enable a interrupt and handled it properly then at the
time of interrupt handler will be called, but how then control reaches
back to u-boot original task that was running prior to interrupt
delivery? I mean who is taking care of context switch at the time of
interrupts?

>
> It's just just most U-Boot drivers do not use interrupts but polling
> mode - note that there are exceptions from this rule, too.
>
> Best regards,
>
> Wolfgang Denk
>
> --
> DENX Software Engineering GmbH,     MD: Wolfgang Denk & Detlev Zundel
> HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
> Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd at denx.de
> A modem is a baudy house.
>

[U-Boot-Users] question about interrupt

[Wolfgang Denk]

In message <7f245da80805200240o7dde7b3bl83875b910d69ad08@mail.gmail.com> you wrote:
>
> So if someone enable a interrupt and handled it properly then at the
> time of interrupt handler will be called, but how then control reaches
> back to u-boot original task that was running prior to interrupt

That's just a normal return from interrupt, like you return from a
normal function call.

> delivery? I mean who is taking care of context switch at the time of
> interrupts?

There is no such thing as a context switch.

Best regards,

Wolfgang Denk

-- 
DENX Software Engineering GmbH,     MD: Wolfgang Denk & Detlev Zundel
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd at denx.de
"Anyone attempting to generate random numbers by deterministic  means
is, of course, living in a state of sin."          - John Von Neumann

[U-Boot-Users] question about interrupt

[Andreas Schweigstill]

Hello!

loody schrieb:
> As far as I know, the cpu will set the PC to irq/fiq vector when the
> these exceptions are triggered. And these pointers are usually at the
> beginning position of bootloader, if bootloader wants to pass these
> functions to kernel, there may be some instructions like "b xxxxxx'.
> 
> If uboot don't handle these exceptions, how the kernel knows these events?

Probably you are confusing a bootloader with something like a BIOS. When
a Linux kernel starts it just copies some data (like information
contained in ATAGS on ARM processors) and activates the MMU. U-Boot just
works when the MMU is disabled because it uses physical addresses.

Usually the interrupt handler code isn't used by *plain* U-Boot, and
on many platforms it may be totally out-dated. The big advantage of this
concept is that you can implement your own interrupt handler without
fearing conflicts with other interrupt handler stuff, so the runtime
behaviour of such a single interrupt handler can be much more
predictable than in complex environments like Linux.

And the answer to the next question which you probably have in mind is:
No, the U-Boot process scheduler or realtime kernel also doesn't use
interrupts because the is no kernel. Everything is based on a main loop.
Yes, on implication is that U-Boot also won't respond to ARP/ICMP ECHO
or similar network requests when it isn't running a network command like
bootp, dhcp or ftp.

And also timer handling is not interrupt driven but instead performed by
polling some hardware timer registers.

Regards
Andreas Schweigstill


-- 
Dipl.-Phys. Andreas Schweigstill
Schweigstill IT | Embedded Systems
Schauenburgerstra?e 116, D-24118 Kiel, Germany
Phone: (+49) 431 5606-435, Fax: (+49) 431 5606-436
Mobile: (+49) 171 6921973, Web: http://www.schweigstill.de/

[U-Boot-Users] question about interrupt

[Wolfgang Denk]

In message <48328D89.9050708@schweigstill.de> you wrote:
> 
> No, the U-Boot process scheduler or realtime kernel also doesn't use
> interrupts because the is no kernel. Everything is based on a main loop.

Actually U-Boot has neither  a  concept  of  processes  or  tasks  or
scheduling or whatever. U-Boot is strictly single-tasking.

Best regards,

Wolfgang Denk

-- 
DENX Software Engineering GmbH,     MD: Wolfgang Denk & Detlev Zundel
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd at denx.de
Wherever you go...There you are.                    - Buckaroo Banzai